An overview of the strategy and vision of how Yahoo! is using semantic web technology for different content and metadata applications. The main focus will be on the internal content platform that was built for creating, managing, describing and enriching content with metadata for building Yahoo! web sites. This will cover issues found during the development in scaling, integrity, explaining the technology and making it work well in a fast moving environment. The presentation will outline how these are used with other web technologies such as search, XML, RSS and Atom in a REST-based web service environment.
Points: Content metadata for description and enrichment Scaling and performance of semantic web technology Semantic web technology value Integration and deployment issues Examples of use
Ontologies are a core building block of the Semantic Web. As ontology-based technology leaves the research lab towards industrial projects, the efforts related to the development and management of ontologies becomes an important issue. The availability of effort estimates prior to the actual investment can make a big difference to the investor. We have developed the ONTOCOM cost model to support project managers to estimate the efforts required to build real-world ontologies. ONTOCOM accounts for a number of factors having an impact on the duration and resources (in terms of person months) related to an ontology engineering project. Worth being mentioned among these factors are the size of the target ontology, the complexity of the business domain, and the expertise of the development team. The model leverages own experiences in building ontologies and a survey of over 40 projects in academia and industry. The talk will:
Explain how the ONTOCOM cost model can be utilized to estimate the effort related to building an ontology Introduce the theoretical background of ONTOCOM Describe a case study from our consulting practice at a globally-acting telecommunications operator
The case study highlights the approach taken by our client to build a large scale ontology for data integration purposes based on industry standards and use ONTOCOM to define an implementation roadmap.
Freebase is a repository for Open Data — built by the community and free for anyone to use. As such it provides a strong foundation for semantic applications, providing an open, common mechanism for entity resolution, relational keys and shared semantics.
This talk will introduce Freebase, its data model and API with an emphasis on how it can serve various roles in semantic applications. Integration of Freebase references in RDF and Linked Open Data models will be covered, as well as ways developers can leverage the relationships between Freebase elements directly in their applications. Methods for handling evolving community semantics and data will be discussed. The talk will target developers and data modelers at all levels.
This session will demonstrate the power of combining semantic technologies with a RESTful architecture. It will demonstrate the application of these technologies on an actual B2B interface. The solution incorporates:
Relational to RDF adapters Federated queries Enterprise ontologies Semantic mediation with OWL/SWRL Semantic queries with SPARQL Resource oriented computing with NetKernal
Semantics, while not in any sense in the mainstream at GE, is acknowledged as a valid research topic. To move to the mainstream, semantics must deliver demonstrable value. In our experience, talking functionality generates excitement; talking technology creates apprehension. Barriers include the lack of concrete evidence of business payback, the immature state of the technology, and the lack of robust, reasonably priced tools that invite non-geeks to play in the game.
Semantics at GE Research is gaining momentum. Our first foray into semantic modeling was in the area of adaptive, model-driven, work-centered decision support user-interfaces. This effort resulted in ACUITy, previously demonstrated at this conference. ACUITy generated interest, if not continued financial support. The ACUITy implementation was largely devoid of the requisite domain models necessary for truly adaptive interfaces. In a more recent project we have developed an environment for creating and applying semantic domain models. The Semantic Application Design Language (SADL) is an integrated development environment for creating models, translating them to OWL/SWRL, and using a reasoner such as Pellet to do consistency checking and processing of DL-safe rules. SADL is aimed at subject matter experts and has quickly gained support in the GE research community, with a more guarded acceptance from the business community. And it is being used for a real business application! SADL builds upon a number of Open Source projects, and is itself being made Open Source. So while we haven't demonstrated that semantic technology is a sound business investment, we are trying hard and it feels like we're getting closer. The synergy between the ACUITy approach to model-driven user interfaces and the SADL approach to domain modeling offers interesting possibilities for the future.
The 'Real Web' of interconnected services, people, projects, documents, and concepts that make up our human experience far surpasses the state of today's World Wide Web. The World Wide Web has proven remarkably adept at interconnecting documents and other virtual resources, but it has struggled to create meaningful links to real-world artifacts. Only by linking to and from the real world can the World Wide Web effectively model the Real Web of relationships. This presentation introduces the new Persistent URL (PURL) service that includes facilities to link the Real Web to the World Wide Web.
PURLs are Web addresses or Uniform Resource Locators (URLs) that act as permanent identifiers in the face of a dynamic and changing Web infrastructure. Instead of resolving directly to Web resources, PURLs provide a level of indirection that allows the underlying Web addresses of resources to change over time without negatively affecting systems that depend on them. This capability provides continuity of references to network resources that may migrate from machine to machine for business, social, or technical reasons. With the recent re-architecture of PURLs to reflect the W3C technical architectural findings for supporting URLs for both network and physical resources, PURLs provide a Web-architectural solution for identifying all resources that relate to business and business processes.
Gleaning Resource Descriptions from Dialects of Language (GRDDL) is one of the newer components to emerge on the Semantic Web Technology stack. Unlike much of the other components that *start* with RDF, GRDDL starts with XML and *ends* with RDF. The combination of the disparity between the development communities bridged by GRDDL, the two representation formats it leverages (XML and RDF), as well as its very mechanical nature, contributes to a significant learning curve for those who wish to understand how it works, what problems it addresses, and where it is currently being used.
This presentation intends to present a full account of GRDDL that focuses more on the "why and where" than the "what and how." In particular, the audience is expected to be comprised of individuals who have never used GRDDL before but are interested in the general kinds of problems it was designed to address as well as how it might be applied to solve their specific problems.
Finally, this presentation will cover how GRDDL mechanisms are being used to build infrastructure for a very versatile patient record system as a demonstration of some of the other lesser known scenarios that this new technology is poised to play a significant role in.
Although traditional work on semantic technologies has involved non-spatial data, a new effort includes consideration of the geospatial component of information. In this way, the semantic technology and geospatial communities are beginning to come together. Geospatial information can be any type of data that describes a geographic location, such as remotely sensed images, vectorfiles, or orthophotography. However, many other types of documents and Web pages can also be geographically referenced to a location. Work in geospatial semantics is needed to take advantage of the geospatial component of information. Such work includes the design and development of geospatial ontologies and tools (both spatial and domain specific) and the Geospatial Semantic Web.
This talk presents projected visions and gives examples of current work in geospatial semantics. Activities of various standards organizations and other groups are included: Vision of Geospatial Semantics Ontologies, technologies needed to achieve the vision, current research, and activities.
Part I: Integrating the Oracle OWL Inference Engine with Complete DL Reasoners The inference engine is a key component of Semantic Web technologies. As Semantic Web technologies mature, OWL ontologies and RDF graphs are growing increasingly in size. Users and application developers of Semantic Web technologies therefore need an efficient and scalable inference engine that is not bound by main-memory size. Traditional RDBMS technology has proven to be a suitable platform for building such an inference engine. In this talk, we first present an architecture for a scalable RDBMS-based inference engine that supports an expressive subset of OWL DL vocabularies. We then explore a simple and yet powerful integration of this RDBMS-based inference engine with external complete DL reasoners, via the Oracle Jena Adaptor. Finally, we demonstrate the scalability of inference performance using benchmark ontologies.
Part II: Oracle 11g Semantic Store: Capabilities, Benchmarks, Tools, Tips, and Best Practices for Building Semantic Applications Efficient and scalable support for RDF/OWL data loading, inferencing, and querying in commercial databases plays a critical role in building semantic applications capable of handling fast operations on large scale semantic data. Semantic technologies support in Oracle 11g adds significant new functionality and provides huge performance improvements.
In this talk we will: Outline the new capabilities including bulk loading, native OWL inferencing, querying, and ontology-assisted querying Provide detailed performance numbers for these features for some of the existing benchmarks Explain how recently supported Oracle Jena adapter allows querying via SPARQL and facilitates efficient integration with tools such as Topbraid Composer Illustrate, via examples, a set of tips and best practices for exploiting the Oracle Semantic Store capabilities for building efficient and scalable semantic applications
Explaining the value of semantic technologies to the public has proven to be a uniquely challenging task: in one sense, semantic technologies will eventually give consumers a much richer experience on the Web, but at the same time, the technology and applications are still in a relatively nascent state. In this panel, a group of industry insiders and analysts will share some of their insights into the challenges and opportunities at hand, including:
What's working and what's not, in terms of educating different markets about the semantic web? What initiatives are connecting with consumers, customers and/or end-users? What does "beta" really mean, and how does that map to expectations in the marketplace? How does the venture capital climate for semantic technologies affect market perception? What needs to happen in order for semantic technologies to make their way meaningfully to market in the broadest possible sense?
Development tools have long been the missing ingredient for companies looking to exploit semantic technologies. Recently however, the tool picture evolved dramatically with the announcement and release of open source mashup technology from MIT.
An introduction to the Semantic Web and all of its intricacies is typically presented by writing a fledgling, consumption-related ontology. Since ontologies do not perform any function in and of themselves, more effort is still required: writing the prerequisite software to reason over the ontology, adding instance data or just comprehending what to do next. This detailed work often dissuades people from experimenting with the Semantic Web. This talk concentrates on jump starting your experience by using existing XHTML/HTML artifacts and learning how to quickly embed semantic-friendly microformats in them. Discussion covers POSH, eRDF, RDFa, GRDDL, XFN and others, in addition to a usable set of tools and methods for getting started immediately.
Nova will report on Twine and the progress of their beta, using the application as a lens to look at user behavior and key learnings thus far. He will also discuss where the service — and the market — are headed in coming months.
The semantic enhancement of Wiki-technology is a major step forward towards the realization of collaborative Semantic Web infrastructures. Implementations such as the semantic extension of the well-known MediaWiki ("Semantic MediaWiki – SMW") enable communities to create, share, and evaluate semantic annotations. The semantic enrichment supports structured queries in addition to fulltext-search, thus helping users to establish "readable knowledge bases." Earlier versions of the SWM-extension suffered from a number of usability issues and a limited technical support for important tasks like visualizing, analyzing, and cleaning the underlying ontologies. Those features are crucial in terms of minimal ontological commitment and the agreement on a common model.
Our goal in the context of the project HALO was to establish Semantic MediaWiki as an easy-to-use platform for domain experts and benefit from its advantages, such as the scalability, the availability as open-source software, and the popularity of the underlying platform. To overcome the limitations mentioned above, we focused on two aspects: the elimination of "productivity bottlenecks" through user-interface enhancements and the systematic support of model cleaning (so called "gardening").
Following a user-centred design approach, we were able to gather immediate feedback from a group of subject-matter experts throughout the development process. Evaluations of the various Semantic MediaWiki extensions, including the analysis of a dedicated "verification-Wiki," have proven the success of this approach.
In our presentation we will describe and demonstrate:
Challenges for collaborative ontology development with the Semantic MediaWiki Interfaces for the efficient editing and maintenance of semantic metadata with SMW Requirements for maintaining a knowledgebase (KB) with SMW Approaches for semi-automatic gardening of a SMW KB Solutions for user guidance and motivation towards a complete and consistent KB
Whales & Cat Fur - Using a Semantic Network to Improve Precision & Recall
Word morphology, parsing and sentence logic are common approaches in semantic methodologies. But adding a language specific semantic network can provide a force multiplier effect in semantic precision and recall. We will explore the construction, content, and relationships among an English language semantic network and then sees its equivalent in Arabic.
Tom Ilube, CEO of Garlik, the semantic web company backed by $21m of venture capital, will present the work that Garlik is doing in the emerging area of "social verification".
One of the big challenges for the next stage of the web will be fighting an emerging wave of ‘identity spam’. An approach to context-sensitive verification that leverages the social graph, what we call ‘social verification’ offers one way of tackling this challenge.
Semantic technologies provide a compelling platform to bring this to life and Garlik will announce an interesting proposition based on its large-scale 60 Billion triple RDF store that points to the potential of this new area of social verification.
The Life Sciences have produced a large diversity of Semantic Data in the form of Ontologies, Taxonomies, Controlled Vocabularies, and other structured datasets to help organize and understand Life at all levels, from large populations down to individual protein interactions inside cells.
Our understanding of Life Science processes is incomplete, making our Semantic resources to have built in uncertainty and error. The fast pace of research in the Life Sciences causes our Semantic resources to continually undergo change.
Constructing a Semantic application in the Life Sciences that provides real value to users is challenged by data uncertainty and multiple sources of error, many sources of Semantic data to be utilized, and continual changes in data versions and formats.
Marc Hadfield will discuss these challenges, solution methodologies, and demonstrate a Life Science Semantic Application utilizing these methods.
Included topics:
* Using Multiple Ontologies in an Application * Handling Differences in Protocols and Formats * Community & Collaboration Processes in Ontology Management * Managing Error and Uncertainty in Semantic Data Sources * Managing Change in Semantic Data Sources * Role Separation in Application Development
Common Logic (CL) is a framework for defining and relating logic-based languages. CL was adopted as an ISO/IEC standard in October 2007. CL defines a common semantics that is upward compatible with commonly used logics and notations such as the Semantic Web languages RDF and OWL, the database language SQL, the UML graphics, and many other notations for knowledge-based and rule-based systems. Common Logic was designed to serve three closely related purposes: to define a foundation for knowledge interchange among semantically similar, but syntactically diverse, notations; to provide a methodology for defining new logic-based dialects for special-purpose subsets of CL; and to support R & D in ontology and knowledge representation.
This tutorial will cover the following topics:
The role of logic and ontology in supporting semantic systems The semantic foundation of Common Logic and its relationship to the logic-base (LBase) of the Semantic Web languages RDF and OWL The use of Common Logic to support interoperability among multiple logic-based notations such as SQL, UML, RDF, OWL, SPARQL, RuleML, and other declarative and rule-based languages The three standard CL dialects: Common Logic Interchange Format (CLIF), Conceptual Graph Interchange Format (CGIF), and the XML-based notation for Common Logic (XCL) Designing controlled natural languages as dialects of Common Logic that can support English-like interfaces to and from semantic systems Issues concerning interoperability, computability, and expressive power of various subsets of Common Logic Features of CL that can support richer semantics in future systems
The ISO/IEC 24707 standard for Common Logic is now available as a [link] target=_blank'>free download for personal.use.
This presentation discusses a framework for the geospatial Semantic Web. Some of the items presented are uses cases, benefits, open source tools, standards, and service orientation. Actual implementations for government customers will be discussed to include implementation and lessons learned.
The SEC is now mandating that XBRL be the business reporting format for public companies. Its adoption will take place over the next five years. In this session, we will discuss this mandate and the challenges the industry faces. Presenters from the financial industry will share their own thoughts about implementations and solutions using XBRL.
This session will discuss how modeling techniques need to adapt when working with the open-world assumption. The traditional modeling assumption is termed 'negation as failure,' meaning that if a piece of information is not present, then it is false. Under an open-world assumption, everything is assumed to be unknown until proven otherwise.
This can be a powerful tool, but it can also trap those who do not adjust for it.
An open-world assumption requires data modelers to model what is not true as well as what is. OWL provides the means to do this, but it requires a change in the way we think about modeling data.
This session will cover:
The basics of the open-world assumption and how it affects modeling practices The axioms in OWL that are relevant to working with open-world data How to 'close the world' when information is known to be complete How to bring closed-world data and open-world data together into a single data model
This talk presents an overview of formal ontology and a business case study of its application. The talk includes how formal ontology compares to less formal approaches and how the Suggested Upper Merged Ontology (SUMO) (www.ontologyportal.org) compares to other formal ontologies.
Classes of ontology-based applications are introduced A detailed description of first order logic is provided Issues of the capabilities and tradeoffs in first order logic inference are explored Several exercises are included in the tutorial in order to maximize audience understanding of the concepts and provide the basics needed for ontology creation. The SUMO is also described in detail, along with its mappings to the WordNet lexicon
The classical view of categorization is that categories are defined in terms of common properties of their members. Prototype theory, pioneered by Eleanor Rosch and her colleagues, proposes a different model, that categories are not so rigidly defined. A category can contain exemplar members ('prototypes') that share more properties with other members, as well as outliers that share fewer properties. Rosch also observed categorization effects dependent upon the level of specificity of the category ('basic level effects'). These concepts inform our approach to designing a database-neutral meta-schema for data organization, for a wide range of information domains. We will review the relevant concepts from Prototype theory and present our implemented architecture for meta-schema based upon these ideas.
We will show how SAWSDL can be used to capture the desired functionality of partner services and how this description can be leveraged during the runtime for discovery and binding. We will cover a set of open technology tools and how they connect to existing industry SOA tools.
Detailed insight into the SAWSDL object model including an overview of the SAWSDL OM, creation and manipulation of the OM using SAWSDL4J Publication and discovery of SAWSDL services using SEMRE, a jUDDI-based Web services registry supporting SAWSDL Run time and selection and binding of partner services using Apache Axis 2 Demonstration of how all the above mentioned technologies work in conjunction with Oracle BPM and Active Endpoints Active BPEL engine
The artifacts used in the tutorial--including the source code, third party libraries, and ontologies--will be available for download at [link]. The participants of the tutorial will also be given a CD copy of the artifacts.
The material assumes some familiarity with SOA technologies such as jUDDI and Apache Axis 2.0, though detailed technical knowledge is not required. Several examples will be presented as Java code fragments, so some familiarity with Java and associated technologies is expected.
The current explosive growth of online video necessitates improved ways to manage and navigate such audiovisual content. This talk discusses the role that semantic computing can play in the editing, publishing, syndication, and discoverability of online videos and presents a concrete, successfuly-deployed application that applies speech, language, and semantic technologies to automatically convert a newscast from its full-length broadcast (long form) to segments containing single stories (web clips) and prepares such clips for online publishing and semantically-aware syndication.
Technologies discussed:
Audio/speech technologies: ** Voice and music detection ** Language, gender, and speaker ID ** Phonetic indexing and search ** Transcript synchronization Language/semantic technologies: ** Document classification ** Named-entity extraction ** Taxonomies and semantic relatedness ** Search query analysis and validation
This tutorial demonstrates how to build practical Semantic Web application using open data and open source semantic technologies. Using a hands-on approach, participants will walk through semantically annotating data, joining disparate data source with non-deterministic result sets and republishing semantically annotated data.
The tutorial is broken into three sections: An introduction to applied semantic techniques Introduces the value of shared semantics between applications. Provide a brief contrast between the formal Semantic Web and current applied approaches. Identifies 'strong semantic keys' and the need for annotated open data as a critical elements shared between these two communities.
Weaving semantic relationships with today's technologies Introduces the basic concepts and representation of RDF data sources. Identifies the mechanisms for reading and navigating Linked Open Data[1]. Describes how to use Freebase as source for shared semantic and strong keys.
Building something useful: a semantic widget for movie reviews A simple blogging widget is described which brings information together from disparate open data sources based on a single, author configuration parameter. The widget performs simple attribute introspection allowing the author to configure the display of the widget based on the data gathered. The author's content is republished with additional semantics making the contribution usable by other authors in the future.
Developers attending this tutorial should feel comfortable making simple REST API service calls in their favorite language and have a passing familiarity with basic XML and JSON constructions. Though code snippets will be presented in Python their translation to any modern language should be obvious.
As background, this tutorial is an effort to demonstrate that the vision painted by Ankolekar et al.'s paper 'The Two Cultures: Mashing up Web 2.0 and the Semantic Web' ([link]) is not only possible using today's semantic technologies but through data reuse and extension these techniques can ignite Open Data Communities.
The workshop will cover essential issues including:
Definition and purpose of a taxonomy Different types of taxonomies Taxonomy design and development Taxonomy integration Automated taxonomy generation Taxonomy governance and management Multilingual taxonomies Taxonomy applications in search, browse, and discovery contexts
This hands-on tutorial will demonstrate how semantic technologies and Resource Oriented Computing (ROC) techniques may be combined to create the next generation of enterprise services. Using NetKernel, an Open Source ROC environment, and Mulgara, an Open Source semantic database, we will show how to build a scalable RDF-processing framework on top of data sources that already exist. This type of integration can be trivially integrated into existing Web and application server infrastructures to semantically-enable more conventional Web-based systems. We introduce the use of Persistent Uniform Resource Locators (PURLs) to relate real-world and Internet-based resources into the World Wide Web's addressing scheme. Introduction to REST, ROC, and PURLs NetKernel as a ROC engine NetKernel for XML processing Extending NetKernel for Semantic Web processing Introduction to GRDDL/RDFa RDF storage and query with Mulgara Example pipelines such as Consumer-oriented: GRDDL: harvesting vcards, ical, RDF/XML Producer-oriented: RDFa: Semantic markup of existing XHTML Consuming RDF in the browser with Project Simile Project Simile Firefox extensions RDFa and AJAX Repairing 404s with PURLs PURLs as the basis for a consistent resource identification strategy
The combination of Structured RESTful services, RDF, and PURLs provide a powerful basis for creating a Resource Oriented Computing platform that can be overlaid on existing organizations and business processes that enable more effective data integration. This talk will discuss these techniques and demonstrate the effectiveness of cross-boundary data sharing leveraging persistent identifiers.
SPARQL is fast becoming the standard query language of the Semantic Web, yet it seems mystical and difficult to many beginners. In this tutorial, we de-mystify the workings of SPARQL with a series of hands-on exercises and examples, starting from very simple graph patterns, and moving up to more elaborate combination queries. 696Each stage is designed to move the learner one step along the path to SPARQL expertise.
A variation of this workshop has been given over the past two years as part of the TopMIND Semantic Web training. We have tuned this module to respond to the most common difficulties and misconceptions that beginning SPARQL users have.
We move beyond SPARQL to describe how to combine queries with popular inferencing paradigms and to control sequences of queries, eventually embedding the results in dynamic web page presentations of semantic data.
SPARQL Graph patterns Comparson to well-known query paradigms (SQL) CONSTRUCT queries Advanced SPARQL: OPTIONAL, FILTER, and negation SPARQL scripting SPARQL-based JSP
This panel includes representatives from large companies who were early adopters of semantic technologies. Come learn from the experiences of Sandia National Labs, Raytheon, Harland Financial, and Eli Lilly, as the panel tackles some of the key questions in getting started with building semantic apps: What is different and what is the same? Where should you start? What skills and training do people need to get started? What is the single most important thing to get started successfully?
Oracle Database has support for native storage, querying, and inference of semantic datasets containing hundreds of millions to billions of triples. This scalable and secure infrastructure can be used to build applications for data integration, knowledge representation, ontology usage and management, and so on. In this workshop we will first give an overview of the technology and describe some customer success stories, and then step through an in-depth technical tutorial followed by a presentation of performance results. The session will end with presentations from our partners describing their integration with Oracle and use cases the integrations are relevant for.
As the Semantic Web continues to gain momentum and as more and more data becomes available, more of the full vision of the Semantic Web is becoming achievable. In this tutorial, we will introduce several mechanisms for accessing, manipulating, and analyzing Semantic Web data, starting with W3C standards and ending with some current active research work. We will discuss approaches to querying and reasoning over this data and introduce participants to some of the tools available. We will mainly focus on: SPARQL, a query language for RDF that can be used to express queries across different data sources N3Logic, a logic that allows rules to be expressed in a Web environment AIR, a policy language that is focused on generating explanations and proofs for security and privacy policies.
We will demonstrate tools for each of these technologies in the context of some specific prototype applications and will provide the opportunity for participants to try these tools on-site. We encourage participants to bring their own laptops.
RDF is a key W3C specification and a foundational component of the Semantic Web. This tutorial will explain the basics of RDF and how it functions as a key building block of semantic systems.
Web 3.0 is the next phase of internet evolution. It uses semantic technologies to cope with challenges of scale, complexity, security, and mobility, as well as provide intelligent, rich media interaction and autonomous behavior that make our experience of internet more relevant, useful, enjoyable (and profitable). This shift from information-centric to knowledge-centric patterns of computing and communication will fuel billion-dollar technology markets and trillion-dollar economic expansions worldwide over the next decade. This tutorial explores the outlook for semantic wave markets over the next decade. We define the semantic wave: it origins, key driving forces, technologies, market directions, and likely developments. We highlight semantic applications, case studies, and adoption scenarios in 14 horizontal and vertical market sectors. We examine the growth of supply and demand for products, services, and solutions based on semantic technologies. Those planning to attend should first download and read the free Semantic Wave 2008 Report — Executive Summary, which is available at: [link] stc2008005
Various treatments of rules have been suggested for different purposes. For example, some treatments (e.g., production rules) feature an "If—Then" syntax. Are these treatments suitable for direct use by business people and analysts? If not, what treatment is appropriate? What makes a treatment of rules ‘semantic’? What are the practical concerns for using rules directly to run the business and to encode and deploy its operational business knowledge?
This presentation starts by examining the treatment of rules under Semantics of Business Vocabulary and Business Rules (SBVR), a standard finalized by OMG in September 2007. SBVR, which is based in semantics and logic, formally defines a general business-oriented classification of guidance including rules. Mr. Ross explains the SBVR categories of guidance, providing relevant examples.
Then, based on real-life experience with developing business rules for large problems, Mr. Ross identifies why alternative treatments of rules, especially using If-Then syntax, break down at scale. He suggests what approaches are needed to effectively capture, express, manage, and deploy literally 1000s of rules in a business.
The elements of business guidance What ‘rule’ means in the real world and why that is important SBVR on business rules Why no ‘action’ in expressing business rules The imperatives of scalability and large-scale re-use Rethinking IT practices for the knowledge economy
Effective communication requires a common vocabulary. An ontology provides a description of the terminology, concepts, and relationships for a particular area of interest. An ontology may be viewed as a declarative encoding of the meaning of the domain vocabulary terms, thus making it a key to enabling communication. For systems that are used by people whose understanding of a domain is not necessarily consistent, an explicit description of the important terms can be extremely useful. Many commercial companies have successfully deployed applications with increasing use of semantics such as taxonomy-based search and navigation services. Rule-based manufacturing, product configuration, and financial services systems have been relatively common in those industries for many years. Fewer organizations have successfully deployed semantically rich systems that incorporate ontology-based metadata, sophisticated reasoning, and explanation support. The technology has been around for decades, though its use for web-based applications is relatively recent, and it remains difficult for some people to understand, let alone use effectively.
This tutorial provides an overview of the knowledge representation landscape and attempts to demystify some of the 'black art' of ontology development. We will outline basic methodology steps developed from a combination of the following:
* Domain analysis methodology from software engineering * IDEF methods developed for the US Department of Defense * Best practices from the Semantic Web (Best Practices and) Deployment working group * Experience and lessons learned Examples from agriculture, finance, healthcare, government, and insurance domains will be used, with a focus on the Web Ontology Language (OWL).
We will also demonstrate appropriate use of OWL DL, OWL Full, and more expressive languages such as Common Logic, to help potential users understand both the power and limitations they impose on applications in making such choices. Insights into some of the changes coming in OWL 1.1 will also be provided. This popular tutorial is back for the third time and provides provides a great introduction for those who are just beginning to 'get their feet wet' in the field.
Service-Oriented Architectures (SOA) are part of a long string of "silver bullet" technologies that are compelling and well-intentioned, but ultimately fail to live up to their promise. The reasons for this failure are beginning to be well-understood, however, and we can salvage the vision with some new uses of established technologies from the Web.
Physical architectures do not focus on a single aspect of the concrete instance; neither should our software. We need to focus on the physical and logical connections between a system's components and the flow of information through it. Additionally, we need to track information about the architecture and its data, map the business goals into the systems that support them and consider how our designs enable or inhibit regulatory compliance. This demanding set of requirements must be balanced by leveraging existing systems, team skills and an eye toward sustainable maintenance.
Semantically-enabled SOAs require the ability to express business concepts in the language of the implementation technologies in ways that are accessible to the non-technical stakeholders. This is a hard challenge to meet, but the Web and its semantic-oriented technologies continue to prove themselves in meeting this challenge. This tutorial will demonstrate the vision, the concepts, the patterns and guidance on implementation roadmaps for successful SOAs by leveraging these technologies. Specifically, we will cover:
addressing your documents, data, services and concepts with consistent, well-designed, long-lived logical names using a consistent metadata strategy for describing the topics of interest to your business use and reuse of existing systems, databases, services and components with minimal effort supporting a transparent architectural migration strategy driven by business, not technical, requirements passing around references to data rather than data itself to enable context-specific access control and logging strategies transforming the structure and level-of-detail of your information as it is being accessed while simultaneously enabling efficient use of computational resources decreasing the cost of and increasing the success rates for supporting regulatory compliance
Social networking sites have captured the attention of millions of users as well as billions of dollars in investment and acquisition. As more social networking services (SNSs) form around the connections between people and their objects of interest, and as these 'object-centered networks' grow bigger and more diverse, more intuitive methods are needed for representing and navigating the content items in these networks: both within and across social networking sites. Also, to better enable user access to multiple sites, interoperability among SNSs is required in terms of both the content objects and the person-to-person networks expressed on each site. This requires representation mechanisms to interconnect people and objects on the Web in an interoperable, extensible way.
The Semantic Web provides such representation mechanisms: it links people and objects to record and represent the heterogeneous ties that bind us to each other. By using agreed-upon Semantic Web formats to describe people, content objects, and the connections that bind them together, SNSs can interoperate by appealing to some common semantics. Developers are already using Semantic Web technologies to augment the ways in which they create, reuse, and link content on social networking and media sites.
In this presentation, we will give an overview of various social networking and social media applications, list some of their strengths and limitations, and describe some applications of Semantic Web technology to address issues with social media sites and to enhance the current 'Web 2.0' platform with semantics. We will demonstrate how the Semantic Web can serve as a useful platform for linking and performing operations on diverse person- and object-related data gathered from heterogeneous social networking sites and show that in the other direction, social media sites can themselves serve as rich data sources for Semantic Web applications.
The history of the Semantic Web goes back several years now. It is worth looking at what has been achieved, where we are, and where we are going. Ivan Herman, Semantic Web Activity Lead for the World Wide Web Consortium (W3C) leads us through this as we prepare for a week of deep discussions with people from all parts of the community of semantic technologies.
This presentation is designed for attendees who are relatively new to Semantic Technology. We will help you get the most out of the conference by rapidly bringing you up to speed on the technology, terminology and concepts that will be discussed in the remainder of the conference.
What is Semantics? What does it have to do with information systems? What is inference? How does it work? The five key subdomains of Semantics:
o Discipline
o Tools
o Content
o Infrastructure
o Standards
Description and Examples of Semantic Technology applied in a variety of settings:
o Knowledge Applications
o Smart Search
o Federated Query
o Enterprise Application Integration
o Unstructured Data and Entity Extraction
OWL and the Semantic Web Basic OWL Concepts An Example Ontology
Service Consumers such as Emergency Operations Centers must find trusted Service Providers quickly in critical situations. For interoperability, both must implement Standards. Different jurisdictions use different vocabularies for event types, organizational roles, and resource names. These differences cause problems if no means for reconciling semantic differences is available. When healthcare's a concern in emergencies, there can be be problems with standards-implementation across applications, operating systems, platforms, and jurisdictions.
A new group, Integrated Response Services Consortium (IRSC) provides these resources for Service Providers and communities in Emergency Management and Healthcare through a Service-Oriented Architecture (SOA). IRSC runs SOA Registry-Repositories (SOA-RR) for Emergency Management and Healthcare.
This SOA-RR provides service-listings for sirens, radios, reverse 911, geospatial location services, hospitals, clinics, laboratories, results-reporting, and decision support.
This presentation covers profiling the OASIS SOA Reference Architecture to the ebXML Reference Information Model (RIM) on Sun's Service Registry.
Semantic Universe and Orbis Technologies, Inc. are pleased to announce plans for a Government Sidebar at SemTech. The Sidebar is open to all attendees of the 2008 SemTech, but will be tailored to attendees that are Federal, State or Local Government employees and government contractors. The purpose of the sidebar will be to discuss government-specific issues in adopting and implementing semantic technologies and networking across government domains to learn about successful implementations and common benefits.
The format of the sidebar will be group discussions with invited expert panels / discussion. Many of the leading semantic technology vendors will be in attendance at the Semantic Technology Conference, which provides the sidebar with industry’s best and brightest to answer our toughest questions, and help us solve our toughest problems.
This 'Fishbowl' is a roundtable conversation amongst audience members on a variety of issues which have arisen during the conference. Panelists will be listed as we get nearer to the event, and topic areas will be posted on the conference notice board during the meeting so you can know what to expect. The format of this session is quite unique and it was one of the best received conversations at last year's SemTech Conference.
Every year, the U.S. Census Bureau publishes the Annual Statistical Abstract, 'the authoritative and comprehensive summary of statistics on the social, political, and economic organization of the United States' as a large set of downloadable Excel spreadsheet files. This government data is not readily accessible to Web search engines and cannot readily be shared, reused, and analyzed in new contexts.
This talk will present joint efforts between Cambridge Semantics, the U.S. EPA, and the Federal Semantic Interoperability Community of Practice (SICoP) to integrate semantic technologies, spreadsheets, and the Web to overcome many of these shortcomings. In particular, by representing information in the Census Bureau's spreadsheets as RDF data backed by definitions in a common semantic repository, shared concepts and relationships between different agencies' data is easily discovered and exploited. And by treating the spreadsheet as a user interface for manipulating semantic data, the data can easily be presented on the Web, where it is automatically updated when the underlying data tables change. This presentation will demonstrate the following in the context of the data that comprises the U.S. Government's Annual Statistical Abstract:
The use of Cambridge Semantics's SHAPE middleware platform to extract semantic information from Microsoft Excel spreadsheets A semantic repository containing shared definitions of data table columns that can be created, extended, and reused via a tightly integrated user interface in Excel Real-time changes to information that are reflected in other spreadsheets Repurposing the spreadsheet-based data tables onto the Web, while maintaining a live connection to the authoritative spreadsheet tables Guided search and query across the data from different spreadsheets
The primary goal of the KMI is to provide Eastern Range stakeholders with a semantically unified, web-based view of distributed range information -- a Single Integrated Range Picture -- through a virtual, federated, ontology-based enterprise model we call the Knowledge Management Framework, which constitutes a 'Semantic Service-Oriented Architecture.'
We present progress made since last year's SemTech07, lessons learned, and technology currently under development to support ontology-driven EII, reasoning, and search. We also discuss the visibility being gained by the KMI success story at high levels in USAF community.
This year, we provide a high level overview of the following progress, along with clear indications of the technical readiness in each technology/application area:
* New technologies and technology providers added to the KMI Team since last year * New applications, both fielded and under development, that access and analyze federated information using our ontology * Growth of the underlying 'Semantic SOA' information * Ontology-based search through unstructured information * Ontology development tools * Agent-based scheduling and a proposed extension to OWL, called Schedule XML, to support schedule knowledge representation * Early work in integrating publish/subscribe services and listeners to the architecture * Reasoning over-federated range data to support decision-makers * Development and deployment methodologies, along with lessons learned along the way * Innovations in funding and making the business case (repeatedly) * Alignment of our approach with the Network-Centric Data Strategy and the USAF Metadata Environment * Hard problems yet to be solved * The way ahead
We conclude with a demonstration of some of the more mature applications being deployed.
Business users and content managers may be ready to create the taxonomy or metadata schema to support a website or Semantic Web application, but what tools should they use to build it? At one end of the spectrum, they could put the whole mess into a spreadsheet and try to wrangle it into some sense of order. At the other end, they're scratching their heads over applications designed for developers. Sometimes it may seem like all the tools available are either too feature-poor, or too complex to be easily usable.
It's time to survey the available tools - commercial and open source - reviewing their strengths and weaknesses. As we find the gaps, we'll also discuss what we want and need these tools to be able to do for us. This will be a very interactive discussion, so if you've ever felt like you wanted more from your taxonomy tools, come share the pains you've experienced and your wish list of features for the ideal taxonomy solution.
Semantic technologies have been around a while and share a great and noble pedigree. This is both their strength and their downfall. Too often, projects falter on the grounds of failed buy-in, failed adaptation to the business, categorical ratholes, funding miscalculations, poor tool choices, and organizational fiefdoms. This talk will make hardcore semantic purists wince, information scientists suicidal, and business sponsors and C-levels hopeful, inspired, empowered, and amused. The presentation provides:
Overview of pitfalls and problems viz: **Cost **Organization **Tools **Approach and roles **Complexity Recommendations, learnings, takeaway priorities, and planning how-tos ... All derived from and including a description of case studies from several industries and organizations
Some very promising recent approaches in B2B integration leverage existing supply-chain-management document standards to enable interorganizational processes. One such industry-driven business document standard is RosettaNet, supported by the global leaders in the ICT industry such as IBM, Microsoft, Oracle, Nokia, Software AG, and Webmethods. RosettaNet is highly successful, but still suffers from some well-known B2B integration problems.
Our presentation motivates the need for explicit semantics in RosettaNet deployments to overcome these issues. The talk presents semantic RosettaNet, an integrated business and process management ontology, embedded in a semantic Service-Oriented Architecture, which overcomes the shortcomings of the existing RosettaNet standard. We also connect our semantic RosettaNet to the recent effort of 'open linked data' and show how to design and reference Web ontologies in a B2B integration setting.
The talk provides the audience deep insight into current best practises for enhancing industry standards with semantics to deal with heterogeneities in Supply Chain Management.
OnLine Analytic Processing (OLAP) is a technique that provides business-level users with an overview of a complex data space. Semantic technologies provide a means for amalgamating information from multiple sources into a single coherent stream. There is a natural match between these two technologies, as OLAP provides powerful tools for understanding federated data.
The University of Texas public health center and TopQuadrant Inc. have produced a prototype ('OPAL') that combines semantic integration with OLAP in which data from seven medical centers in the Houston area are combined, using Semantic Web technology, and processed and displayed using standard OLAP tools.
This presentation will outline the following:
Business value for public health management of a combined Semantic-OLAP capability Organizational challenges faced by the combination of these technologies Technical challenges Demonstration
The Semantic Web is ideally placed as a technology enabler in the coming connectivity disruption. This wave of disruption is being driven by powerful changes in networked social behaviour, the technology of Internet-scale applications and the economics of the ecosystems around them.
The current teenage generation is growing up with the ubiquitous, always on, socially networked Web. As this generation enters higher education and the workplace, they take with them a radically different world view to the previous generation. They understand that we achieve more through sharing and collaboration than from being closed and isolated.
At the same time, the World Wide Web is enabling a radical transformation from islands of content and data to densely interconnected information and data spaces, bringing huge value to those organisations with the scale and algorithms to exploit the value in the connections.
Open source and the ubiquity of Internet connectivity are changing the economics of doing business enabling new collaborative models between organisations and individuals. Innovation networks and open business models are becoming a reality, requiring exchange and integration of unprecedented amounts of data, a key strength of Semantic Web technologies.
This presentation will examine these trends and analyse the global commercial opportunity that the Semantic Web presents in the context of the evolution of the World Wide Web and Semantic Technology.
The Semantic Web enables a new breed of applications that manage disparate knowledge from many sources. These applications need to be able to integrate many different types of data and be tolerant of imprecise and incomplete information. Traditionally development of Semantic Web applications have required significant investment of time and effort in building core scalable infrastructure, diverting effort away from building true user value into the application.
The Talis Platform enables rapid development and deployment of rich Semantic Web applications that connect and integrate data across organisations and the World Wide Web. The Talis Platform provides solid infrastructure for the Semantic Web and is being used right now by Talis and organisations in our developer program to deliver innovative new applications.
In this hands-on tutorial we will provide a step by step guide to building applications using a range of Semantic Web technologies including RDF and SPARQL. We will develop a full and usable application in PHP to demonstrate the ease and accessibility of modern Semantic Web development, backed by a robust platform infrastructure. During the tutorial we will cover modelling the problem domain in RDF; designing the application information architecture; using Web Services to interact with the Talis Platform; searching and browsing by taxonomy; strategies for handling unexpected data and integration with other Semantic Web applications.
The objective of this presentation is to provide technologists, program managers, and buyers of semantic technology products with an overview of the entity extraction market. This presentation serves as a starting point in their due diligence process in determining the value of the technology in this market and the available vendors/products.
The Entity Extraction market segment is often a starting point in deploying semantic technologies. The technologies in the market segment provide the capability to extract information from text. Some of the products/vendors in this market segment use standards, such as RDF/OWL, to represent extracted information.
The author is currently working on large scale extraction projects for federal government customers, specifically tactical intelligence programs. This presentation provides a summary of this market segment: Scope and size Technology and financial drivers Trends shaping the market Considerations prior to purchasing Market leadership valuation and profiles Lessons learned
Robust knowledge representation schemes, a rich and expansive ontology, natural language understanding and generation capabilities, and sophisticated inference mechanisms make Cyc a valuable asset for developing a wide range of powerful semantic applications. While some applications need only use light-weight access to Cyc’s knowledge base via its Java API or web services, other solutions benefit from tighter integration with Cyc’s question-answering, semantic data integration, information management, and explanation generation functionality. This workshop will present a sampler of Cyc applications that demonstrate the value that can be achieved at various levels of integration, ranging from a simple web-services mash-up through a bespoke application developed, with Cycorp support, using the Cyc Analytic Environment platform. Session participants should achieve a solid understanding of how Cyc might be used to semantically enrich their current or planned applications, and what level of development, skill sets, and integration effort would be required to achieve those benefits.
Join us onsite at Oracle’s Customer Visit Center for a three hour deep dive into Oracle’s vision for Semantic Technology. As the industry leader in Grid computing, Business Intelligence, Database and Service-Oriented Architecture software, you’ll be interested in learning about how Oracle is applying semantics to deploy next generation tools in these areas. During this visit you’ll learn how Oracle views the business drivers for the technology uptake from an Enterprise software perspective and what Oracle is doing to roll-out the technology in real-world scenarios. To conclude the visit, Oracle will host a talk-back session where attendees will have a chance to ask questions of Oracle management about the technology, products, or adoption strategies.
This field trip was scheduled in conjunction with SemTech 2007 and received a very favorable response from the people who took part.
Note: Transportation to and from the Oracle campus is provided to attendees.
In this workshop, students will construct a Semantic Mash-up application from end to end, starting with data capture and analysis, data merging, semantic modeling, display and finally finishing with deployment. Each step of the process is fully grounded in semantic technology standards (RDF, RDFS, OWL, SWRL, and SPARQL) and supported by the TopBraid tool suite. Each student will pursue a data mash-up project throughout the tutorial, using data either of their own choosing or from the instruction materials. Due to limited time, the tutorial will be run on a 'just enough' basis; 'just enough' of each technology will be presented to handle the day's application. The deployed system will be built entirely using the principles of Ontology-driven software development; all customization and construction of the system will be achieved using modeling constructs in RDFS, OWL or SWRL. Students will have exposure to the following components of semantic deployment:
• Data sources for semantic mash-ups; microformats, RDFA, RSS, etc. • Basic modeling in RDFS and OWL and how it supports data mash-ups • Advanced modeling using SWRL and SPARQL • Display of information using ontology-driven plug-ins • Deployment of a semantic application using TopBraid Live
With a panel of leaders from the semantic technology industry, this session will give us the opportunity to reflect on the many discussions that have taken place during the week of SemTech 2008 and help us map the course as we prepare to extend those conversations back into our workplaces. We will touch on issues of ROI, making the case for semantic technologies in the enterprise, and what to expect in the coming year in the semantic tech space.
The term "Linked Data" was first used by Tim Berners-Lee to describe the idea that data on the web can be connected to other data on the web; in much the same way (conceptually) that documents on the web today are connected to each other by hyperlinks. The objective of Linked Data is to enable the publication and sharing of data openly on the web, in the same way as documents are published and shared openly today.
This panel will describe the status of Linked Data, including current applications and examples, technical standards, adoption issues and future initiatives.
To fully exploit the reasoning power of the Semantic Web sooner rather than later, subject matter experts need to be able to record their knowledge in as natural and familiar a way as possible--in the way they already think about their knowledge (i.e. natural language). The OMG's new Semantics of Business Vocabulary and Business Rules (SBVR) standard provides the connection between vocabularies, definitions, natural language grammar, and formal logic. SBVR Structured English, a simplified version of English grammar, provides a natural language that SBVR tools can understand and have the potential to use to generate OWL models.
In this session you will learn about: The features of SBVR that offer the possibility of generating OWL models The framework for mapping SBVR to formal logic SBVR interpretation in ISO 24707 Common Logic SBVR mappings and transforms to OWL Tool vendors planning to implement this capability
The Healthcare / Life Sciences Sidebar at the Semantic Technology Conference is a moderated discussion geared to:
Provide insight and discussion of niche tools and applications for healthcare, pharmaceutical, biotech, and life sciences related purposes; Investigate possible areas of cross-utilization and application, collaboration, and cooperation between customers, developers and researchers; Enable networking opportunities for healthcare / life sciences oriented members of the community who are interested in semantic technologies.
The session will have a moderator and discussion leaders to get things started, but it is designed to be an open community discussion
This presentation will report on case studies from three client sites where an Enterprise Ontology was built with the express intent of serving as a basis for the development of the messages for the Service Oriented Architecture. Each Enterprise Ontology was based on the upper ontology gist.
This presentation will cover techniques uncovered to convert existing schema to OWL, how the upper ontology was used to disambiguate concepts in the existing systems, and approaches used to generate XSD from the ontology.
Semantic Web standards are designed with the data integration in mind. Key integration capabilities foundational to RDF include globally unique identifiers and the ability to merge information. Data integration is of high importance in life sciences where a wealth of different data sets contain separate, but related, information about drug trials, chemical decompositions, etc. Integration can produce important new insights, for example, a drug failed in a trial may be discovered to have applicability in another area. As a result, interest in semantic technologies is high.
While standards are important in supporting integrations goals, tool capabilities are equally critical. In this case study, we will present a journey of trial and error in applying Semantic Web technology to the integration of life sciences data. A variety of database technologies and RDF providers were tested as part of the proof of concept (including TopBraid Suite, AllegroGraph, and Sesame).
In this talk we will:
Describe the steps taken and products used Highlight successes and failures and identify missing capabilities Discuss insights on the role of RDF data and models
The Norwegian Defence is faced with challenging information integration tasks and new information sharing and information availability needs -- both between Defence Services and Agencies as well as coalition partners and non-governmental organisations.
At the same time, semantic information integration (SII) is emerging as a better way to integrate information from heterogeneous data sources. By creating a logical hub-spoke structure between ontologies at the knowledge layer, SII can potentially reduce the amount of integrations necessary to combine data from existing systems.
The motivation for our experiment was to explore the use of SII to translate between different XML formats. In the presentation, we show how we used SII to translate between two formats for sharing military situational information at the NATO Coalition Warrior Interoperability Demonstration (CWID) 2007. For the implementation, we used different mapping mechanisms like OWL, SPARQL CONSTRUCT, and Jena rules between ontologies together with hybrid reasoning to implement the format translation.
Programmers writing the semantically-driven applications of the future will need to do more than simply encode their data in RDF and run queries. Object-oriented languages, such as Java, work with classes that often represent real-world entities. These classes, and the classes that process them, can integrate better with other systems and allow more sophisticated reuse if each entity in the system has appropriate semantic markers.
Semantic code markup has an additional advantage. Open source projects will continue to play an important part in Java development. Programmers will be better able to find libraries that meet their needs, if the code is marked so that input/output types and processes can be discovered by Semantic Web searches. This workshop explores the following:
Semantics of Java code and classes Annotations and semantic markup Open source libraries Semantic source code searches Integration with other applications Integration with the Semantic Web
Using RDF/OWL for reasoning about real world problems requires an understanding of the RDF/OWL axioms (which encode the formal constraints of the language) as well as tools for parsing, validating, and processing RDF/OWL documents. In this presentation, we demonstrate through practical examples, how this can be achieved with technologies available today. We will describe a subset of RDF/OWL axioms that can be efficiently implemented in forward chaining inference engines such as BaseVISor, Jess, and OWLM and show how to break out of some of the limitations of RDF/OWL (e.g. binary predicates, limited property value constraints) in a principled and effective manner. Concrete examples will be drawn form multiple domains and used to illustrate the espoused techniques.
The National Information Exchange Model (NIEM) is at the forefront of the national agenda for information sharing. Launched in 2005, NIEM is an XML-based data standard that provides a common vocabulary of agreed-upon terms, definitions, and formats for enabling information exchanges within and across the justice, public safety, emergency and disaster management, intelligence, and homeland security domains. However, being reflective of the diverse and intricate nature of its underlying domains, NIEM is large and complex and the vocabulary of concepts defined by NIEM and the methods of representing those concepts are different from what many organizations are familiar with. This can pose significant challenges to the ability of organizations to cost-effectively map their existing data schemas to NIEM. In this session we will describe a semantics-driven approach to semi-automating the NIEM schema mapping process that leverages both ontological and lexical methods. Topics to be covered include:
The ontological characterization of NIEM Using an upper ontology to normalize data schema semantics WordNet-based semantic similarity of ontologically compatible schema components Specifying semantically-derived schema mappings with XQuery for visualization and refinement in a graphical mapping environment Runtime exploitation of XQuery mappings to generate NIEM information exchange packages (IEPs)
Semantic Web technologies were designed from the ground up for disparate information integration. Disparate data and its impact on data quality and consistency remain one of the most difficult challenges facing large enterprises today. Yet, there has been relatively little use of Semantic Web technologies in these organizations.
The adoption of new technology by large enterprises has always been challenging. Recently, SOA and Web service technologies have had some success in making inroads to large enterprises, mainly due to their seemingly direct addressing of business process automation problems, which business executives could readily appreciate. In contrast, the Semantic Web has been nicknamed the "Pedantic Web" due to its frequently esoteric explanations from academics, hindering its adoption.
What is required is the ability to teach business leaders the value of the technology in terms that they can understand. Our panel includes technical leaders that have taken on this challenge and have succeeded in getting these somewhat trailing edge enterprises to embrace the technology. They now wish to share that experience. Topics covered in the panel include:
What advantages of using Semantic Technologies over existing data integration tools such as ETL and EAI that really resonated with business leaders? What objections to the technology has the panel experienced? How did the panel address these objections? How did members of the panel first introduce the technology? How did members of the panel socialize the use of ontologies without it turning into and explanation of the "Pedantic Web"? How did the panel address how Semantic Technologies fit in to the SOA strategies for the enterprise?
The amount of semantically annotated information available on the Internet is constantly growing. Tagging is one of the most popular approaches to annotate and categorize content on the Internet. There are many other techniques that involve hierarchical structures, such as taxonomies. The design of user interfaces for searching or browsing semantic data is an important issue.
Although there were some efforts to propose improvements of the visualizations of tag multi-sets in social applications, not many evaluation studies have been published. In this presentation, we discuss the results of usability studies of chosen interfaces for rendering semantic data. From the perspective of social annotations (tagging), we compare the usability of various customizations of Tag Clouds and Treemaps. We also discuss the results of an evaluation of common user interfaces for hierarchical data compared to a new approach: the Atom interface. On the basis of these evaluation studies, we propose general guidelines for the design of interfaces for rendering semantic data.
We will present: General overview of approaches and problems with rendering semantic datasets Experimental study on usability of techniques for rendering tag multisets (Tag Clouds, TagTreeMaps) Results of usability study of approaches for rendering large and hierarchical datasets (Atom Interface) General guidelines for building and evaluating user interfaces for semantic datasets
Semantic Interoperability Services in a NATO Network Enabled Capability (NNEC) environment defines the scope of this Allied Command Transformation and NATO C3 Agency research project. This presentation provides the results of this R&D, to include prototyping and specification of Semantic Interoperability Mediation Services (SIMS) and Semantic Interoperability Discovery Services (SIDS) that have evolved from various deliverables during the 2005 to 2008 program.
The mediation services (SIMS) focus is on the conversion, context-based aggregation, and the routing of information and consists of a set of modules that can also be considered services:
Semantic Federated Search Services Inference/Reasoning Services Information/Knowledge Store (Registry Services)
The discovery services (SIDS) aims to provide data, information and service asset discovery, content management, and ontology management support for various communities of interest and includes specification of a dynamic metadata repository. Two main services are addressed through discovery service R&D:
NATO metadata Registry and Repository (NMRR) services and Service Discovery Services (SDS)
In any large project there are multiple teams involved, each with their own priorities and requirements. This talk will focus on one solution to integrating diverse domain rules and concepts using OWL ontologies as the pivotal focus.
Many of the properties of OWL class are taken for granted in RDF, but are difficult to apply to a Java domain model. In this talk, we will explore the reusable and mixable properties of ontologies and then demonstrate how these same properties can be applied to Java classes and methods using AOP.
The examples in this talk use Sesame RDF repository store and the Elmo library to map Java classes to OWL concepts.
Today, people want to be able to query a span of information in a user-friendly method and be provided with smart, contextually-relevant results. Many applications provide user interfaces to query for information and tend to be satisfactory to meet the application's requirements. However, these interfaces can be improved with the use of semantic technologies to allow for advance query construction and therefore facilitate an improved answer to the query. The key to improving these query techniques is the use of an ontology. An ontology allows the data to be organized in a way that gives 'meaning' to the structure, thus allowing for a better approach to support semantic integration of disparate sources.
In this presentation, we will demonstrate our approach (i.e. process) in creating a Signal Intelligence (SIGINT) Ontology and show how this ontology was utilized to support semantic integration of many disparate sources within a distributed intelligent agent architecture (used as a type of a service-oriented architecture). We will show how the ontology definition is transformed into a representation in OWL and Common Logic/Java Objects for machine-processing. We will cover the techniques that were used to map legacy databases into the SIGINT Ontology and discuss how information is brought together from many legacy databases to provide a comprehensive result. We will also address the lessons learned as well as the key challenges of our approach.
This presentation reviews the real-world use case of aXiom, a U.S. Space and Naval Warfare Systems Center (SPAWAR)-led initiative to support Web-centric dissemination and analysis of real- and near-real-time data. The project deploys applied semantics and ontologies to establish a unique context management framework and attain new levels of enterprise-class Semantic Services Oriented Architecture (SSOA).
Combining commercial and open source software, the aXiom team designed a dynamic operating environment utilizing ontologies, inference engines, and rules engines to create a comprehensive model of today's complex enterprise. The project uses ontologies to model enterprise constructs and provide semantic services for data mapping and manipulation, metadata storage, conceptual search, semantic inference, and integration with correlation engines. The architecture manages these capabilities within layers of policy and lineage to support governance objectives, information security, and system evolution. This case study demonstrates how ontologies and semantics can be used to deliver the right content and services to the right users at the right time and includes:
A live demonstration of the application A comprehensive review of the underlying architecture Samples of ontologies, including unique lifting/lowering ontologies Explanation of the semantic rules used in the project Use cases for inference engines
Millions of sensors around the globe currently collect avalanches of data about our world. The rapid development and deployment of sensor technology is intensifying the existing problem of too much data and not enough knowledge. With a view to alleviating this glut, we propose that sensor data be annotated with semantic metadata to provide contextual information. In particular, we present an approach to annotating various media on the Web and sensor streams (e.g., video sensor data) with spatial, temporal, and thematic semantic metadata. This technique builds on current standardization efforts within the W3C and Open Geospatial Consortium (OGC) and extends them with Semantic Web technologies to provide enhanced descriptions and access to video sensor data.
This talk will cover:
An overview of Sensor Web and Semantic Sensor Web Enabling techniques and specifications including semantic extension of Sensor Markup Language, ontology driven spatio-temporal-thematic metadata extraction, and extend SPARQL queries Demonstration of a prototype Semantic Sensor Web application that shows a semantic mashup of YouTube and other video content with Google Maps and other services A discussion of how our framework could play an important role in emergent applications, including video on mobile devices running Android OS
Almost 50 million Semantic Web documents are available on the Web in RDF format and can be readily used in applications or websites. Similarly, exposing data as Semantic Web data has an unprecedented potential for advertising products and services.
In this presentation, I will specifically show how:
To build applications and websites which locate and reuse over 30 million Semantic Web data sources thanks to openly available APIs of Sindice To publish Semantic Web data in a way which is both efficient and guarantees maximum precision and recall when harvested by clients or Semantic Web-enabled search engines thanks to the Semantic Sitemap Extension To remix, process, filter, adapt, and produce Semantic Web Data using Semantic Web Pipes—SW Pipes are a paradigm for data processing which can be used standalone or embedded into software, solving a great number of semantic data information processing needs with unprecedented ease
The presentation will be largely demo driven, with several third-party applications being shown both publishing and reusing data, leveraging the above technologies.
Once considered competing technologies, there is an increasing realization that Semantic Web technology is well positioned to address some of the limitations in Web 2.0 frameworks and applications, and vice versa: the Semantic Web community is taking lessons from the vast social and commercial adoption of Web 2.0. The technological convergence between the two approaches toward semantics on the Web is one of the most exciting developments of today. We will discuss the emergence and evolution of this trend, and look at some of the technological solutions that aim to bridge the gap between the two worlds (RDFa, GRDDL, hGRDDL etc.) Last, we examine in particular what this convergence means for Information Retrieval on the Web.
Semantic Web and Web 2.0: a brief history The complementarity of Semantic Web and Web 2.0 technologies Bridging technologies Information Retrieval in a semantic world
There are currently several different approaches to semantics and the Semantic Web floating around. While the uptake of these technologies are going extremely well, there is still confusion about what sort of technology fits where and how it works. The confusion is made worse because the term 'ontology' is used in several different ways. In this talk, I will describe how these different sorts of models can be used to link data in different ways. I will particularly explore different kinds of semantic applications, from Enterprise Data Integration to Web 3.0 startups, and the different kinds of techniques needed for these different approaches.
Citizens as well as businesses demand that governments become more transparent and provide services tailored to their individual situations. Dealing with public authorities is not always easy. Citizens often find themselves confronted with a large number of different institutions, laws, permits, and websites each offering only a part of the information and services they are looking for, and citizens often do not know who to address with their specific problems or questions.
Improving service to the public is an important priority for the Dutch government. Semantic technologies are used to radically improve the customer service. The government is developing a semantic infrastructure to provide the public with context-specific access to all government information and services supporting both laymen and experts in finding and understanding answers to their questions. For this purpose Be Informed's advanced knowledge instruments and electronic forms are integrated with off-the-shelf search technology to support all the front offices and digital processes of the government.
Each company executive/representative will give a status review of where they are, provide a peek at where they are going, and describe why they have chosen their particular breed of technologies and target markets--and why their approach is a best practice.
Status report & live demo Future survey (may include a demo/video/slide show) Core technologies comprising each company's approach Target markets & competitive positioning Debate: Each company representative will state their position on why they're number one
This session is a 'take off the gloves' panel, with each company going head-to-head with their competitors. The moderator will mix things up, possibly bringing in surprise panelists during the debate part of this panel session.
The objective of this panel is to get a realistic perspective on where the commercial leaders are, where they are going, and if they can get there--and by when.
One of the basic principles of the Semantic Web is that knowledge is distributed over a large number of sources. One common pattern of distribution in the enterprise is that people partition their data in different triple stores according to the importance and life cycle of triples. For example, an application might have one or more triple stores for the hard facts, a triple store for inferred triples, a triple store for the provenance of the triples, a separate triple store for the ontologies, and possibly even a triple store for deleted triples.
The 'hard-fact' or 'asserted' triples are 'expensive' in the sense that it is hard to recreate them if they are lost. They might be unique statements made by analysts, or they may originate from costly document analysis processes. Inferred triples are 'cheap' in the sense that you can always regenerate them from the asserted triples. The ontology triples are in a separate store because they might change very often over time, and it doesn't make sense to include them with the hard facts.
This presentation will demonstrate a flexible federated architecture that allows you to create 'federated stores' on the fly that allow for transparent RDFS++ reasoning and SPARQL over these federated stores. This federated architecture will work for stores that are on the same machine, stores in the same data warehouse, or stores that are geographically dispersed.
This presentation will highlight empirical data showing the speed of RDFS++ reasoning and SPARQL as a function of the number of partitions and the 'distributedness' of the databases.
With the emergence of the Semantic Web, the ontology development process has undergone a radical shift: Ontologies are no longer developed by small groups of people, but rather by larger communities spread around the world. This new paradigm brings out a new set of issues. First, we must address the technical issues of simultaneous access to ontologies by multiple users. Second, an ontology-development environment must address several collaboration and social aspects, such as workflows, provenance, discussion threads, change tracking, user rights, ontology views, trust, and so on.
We present Collaborative Protégé -- a standalone and web-based extension to the Protégé ontology editing environment that supports collaboration and addresses some of the above aspects. In Collaborative Protégé, users can discuss and comment on specific ontology components and on changes in the ontology. The tool enables users to reach agreement on modeling solutions by providing support for change proposals and voting. Other collaborative aids are the change tracking of ontology components, the searching and filtering of the comments, and live chat with other users. We will demonstrate the use of Collaborative Protégé both as a desktop application as well as a web client.
In this presentation we will cover:
* Different aspects and challenges of collaborative ontology development * Use cases for collaborative development * A demonstration of the use of Collaborative Protégé by a group to develop a common ontology * The interface to use collaborative features of Protégé in your own application
The amount of information is growing dramatically and quickly nowadays. It is a very difficult time for teachers to keep up with the changes; this problem also affects eLearning. Without constant maintenance, electronic courses rapidly become outdated. Moreover, all of the current solutions seem to underestimate the potential of Web 2.0 architecture. The contemporary eLearning systems deliver predefined, rigid courses which usually do not take into account that users are different from each other.
We would like to present Didaskon, a framework for automating the composition of a learning path for a student. The selection and workflow scheduling of learning objects is based mainly on their description and the semantically annotated specification of the user profile. Didaskon also derives information aggregated by IKHarvester (informal knowledge harvester) from various Social Semantic Information Sources (i.e. wikis, blogs, fora, and other informal sources).
We will present:
Semantics harvesting for eLearning Integration of informal and formal knowledge Profiling courses for a user’s experience improvement Service oriented architecture of eLearning frameworks utilizing semantics
This panel features prominent VCs, some of whom have made substantial recent investments in semantic technology companies. They will offer a variety of perspectives spanning both the opportunities and pitfalls they see in the emerging marketspace.
This session is particularly appropriate for CTOs developing new products and services based on semantic technologies, as well as entrepreneurs trying to understand where the money folks see the best business opportunities.
CYC introduced the process of large-scale ontological engineering in 1984. We learned a large number of useful lessons, the hard way, during those 24 years -- and 1000 person-years --- that we've been building CYC. Not that we generally could've/should've known better in advance, but half of that effort was, in hindsight, wrong. Errors in representation, in methodology, in inference, in scale. Traps we fell into, decisions we had to back out of, half-finished off-ramps to nowhere. Come hear about mistakes that we made, that others might still be making. You'll pay for the whole seat, but you'll only use the edge!
TBA
Pragati's flagship product, Expozé, is an integrated suite of modules that presents cognitively useful perspectives of software systems. It takes a high-level, domain-and-representation-independent cognitive-based approach for solving the interoperability problem for ontology builders, knowledge engineers, and system integrators. Expozé's unique clustering-based technology:
* Finds 'collaboration groups,' (i.e. sets of terms that share the same operational context) * Detects partial or fuzzy relationships between terms that exist within or across systems * Reveals potential mapping regions across concept sets beyond those exposed by using substring matching and stemming * Extracts recurring usage patterns that can serve as useful shortcuts for rapid reuse and componentization * Discovers easily overlooked infelicitous knowledge entry patterns and incompletely specified knowledge * Applies to a wide variety of representations of knowledge such as, knowledge bases, ontologies, schemas, databases, and stylized natural language text * Does not depend on the domain of analysis
Biomedical vocabularies and terminologies (e.g. SNOMED-CT, UMLS) are getting increasing attention in the Semantic Web community due to their applicability in assisting novel clinical, healthcare, and pharmaceutical research. However, more often the biomedical vocabularies are represented in multiple languages/models and comprised of a large number of concepts and assertions. Consequently, the ability to (i) query, (ii) author/review, and (iii) apply reasoning techniques emerge as major obstacles in the wide-scale adoption of such vocabularies. The National Center for Biomedical Ontologies (NCBO) is one of the National Centers for Biomedical Computing (NCBC), providing innovative technology and methods to create, disseminate, and manage biomedical information and knowledge. As contributors to NCBO, we introduce the Lex* framework for overcoming some of these limitations. In this talk, we will present:
LexGrid: an approach for storing and querying vocabularies represented in multiple languages and models LexWiki: an approach for collaborative vocabulary development and management LexInfer: an approach for modular-based reasoning of vocabularies
The Lex* framework is released as an open-source software and can be accessed from [link].
Automatic generation of semantic metadata is highly desirable for libraries of unstructured text. To date, building the natural-language processing technology required to perform this task has been difficult and expensive. TextTrainer, a software application resulting from 10 years of applied research, learns its natural-language processing rules from example input/output pairs, which makes this task easy and cost-effective. Another advantage of learning from examples is that it can learn to process any language (e.g. Chinese) or genre (e.g. legalese). Northrop Grumman's customers have used TextTrainer to generate semantic metadata for several important applications. We present a real-world example of weekly maritime piracy reports being converted automatically to RDF to feed powerful ontological analytic tools.
In this presentation we introduce ODASE (tm) an ontology-driven development approach and platform, which cleanly separates the Business Domain knowledge (at the description level) from the Software Engineering knowledge (at the execution level). The process for transferring the Business knowledge from the ontology to the programming language is by automatic generation of source code. The power of ODASE is that the model specification, the code generation, and the runtime reasoning use the same formal description.
We present how we used ODASE to build a business-critical e-insurance project for a multinational insurance firm, where only 35% of the requirements were known at kickoff. We required one third of the time of the next closest quote for the project, and a similar project built classically at another insurance firm required also around three times the resources.
We describe an innovative, semi-autonomous Knowledge Extraction capability (KNEXT) that used techniques from text-mining, natural language processing, and knowledge representation. A key design goal is to strike the correct balance between a completely manual knowledge creation process and a fully-automated knowledge acquisition system. Initial applications of this capability suggest that this goal is achievable by clearly identifying those aspects of the process that should be handled computationally and those parts of the process that should be handled manually by the knowledge engineer. Our experimental investigations on several large scale applications indicate progressive efficiency gains for the knowledge engineer as the ontology acquisition process advances. We will describe the results of using KNEXT on Knowledge Management Initiative (KMI) applications at the Air Force 45th Space Wing. We show, through several real examples, how KNEXT is being used to cost effectively build knowledge models for multiple KMI semantic technology applications.
Contextual advertising applications can leverage a number of semantic techniques that infer the relevant topics and the 'aboutness' of a document or webpage. This talk describes the underlying architecture of a new Adobe semantic analysis platform and how it is used by 'Ads for Adobe PDF,' a recently launched opt-in service that enables commercial publishers to monetize their PDF content through contextual advertising.
The talk will discuss how several hybrid semantic analysis techniques were developed in order to build a system that can generate a semantic essence for any submitted document. Using this generated document essence, ads from an advertiser pool provided by Yahoo! are then automatically matched and displayed as dynamic, contextual ads in a panel within Adobe Reader and Adobe Acrobat when a user views the PDF. This talk will focus on how PDF content is systematically analyzed using analysis algorithms that take into account several factors.
Topics in this talk include:
Use of new keyword and topic analysis methods, subject distribution, and TF-IDF measures Discovering topics or themes using general knowledge taxonomies and thesauri Generalization of topics to associate more useful categories and concepts with the document; for instance, content about soccer is generalized and associated with the concept of sports Identification and removal of semantic noise that can incorrectly bias the aboutness of a document Identification of keywords and topics that incorrectly trigger sensitivity filtering by the ad provider Future trends and challenges for contextual advertising systems
Rearden Commerce is a company founded on the vision and power of a Semantic Web. Rearden Commerce provides to more than 1,000 customers an online services marketplace, including air, hotel and car reservations, Web and audio conferencing, and restaurant and entertainment booking. The benefit Rearden Commerce brings to its customers is the ability of the Rearden Commerce online Personal Assistant to intelligently search more than 137,000 merchants in its network and find the most relevant result based on each user's personal preferences, previous searches, current location, and spending policies. Leveraging Web 3.0, a customer's search returns the most accurate, intuitive result rather than a catalog of numerous options to sift through and choose from.
This session is a case study of how Rearden Commerce's customers are using the Rearden Commerce Personal Assistant in the enterprise and how Rearden Commerce is advancing the Semantic Web to bring human intelligence to the Internet.
We all know that much information is spread through the web, and a lot of time is wasted by aggregating this into useful content for better analysis and decision making support. For many applications like BI or CRM, there is also a need to consider 'qualitative' information, which is available in special internet sources like social networks or as free text in news articles. The problem is therefore how to optimally integrate quantitative information from the open sources and the day to day information provided by specialized news agencies.
Ontos developed a new generation of modules founded on semantic based knowledge and content systems. The Ontos Semantic Framework integrates Semantic Web and Natural Language technologies to enhance knowledge acquisition. The key task is to automatically aggregate, merge, and enhance information to meaningful business information. The vertical solution is available for business, economics, and politics using the appropriate ontology for information extraction. Just imagine you have to visit a customer and by one click you get inside your intelligent portal all aggregated information without a lengthy search of the internet.
Object-oriented design is the preferred paradigm for modeling complex software; however, most RDF APIs are tuple-oriented and lack some of the fundamental concepts in object-oriented programming. This talk introduces the Elmo library and how it can be used to model objects in RDF using object-oriented designs.
Topics covered in this talk include:
Basic concept mapping in Java, showing a simplified object/property interface Assigning behaviour and responsibility to RDF resources
The Sesame RDF repository will be used for all examples, and most of the examples used in this talk will be in Java, but other languages such as Ruby, Groovy, and JavaScript will also be demonstrated running on top of the JVM.
The conventional Search Engine Optimization (SEO) strategy, which largely addresses the structure (indexing and tagging) of a website, and not its core content, neglects a large part of what drives search rankings. This is because Google and other search engines have started using a "landing page quality score" to determine the final ranking of a page. This quality score is determined by analysis of the prose text itself, rather than the external links to that page, or its tags. We find that semantic tools which serve to clarify the text of a page (or even enrich it) can render the page’s quality more visible to contemporary search ranking algorithms. For sites having substance, this results in a big boost to SEO. In discussing our recent case studies, we list several tactics that don’t work (e.g. simply spraying semantic terms on the side bar of a page) as well as the trade–offs involved in using semantics for large–scale SEO (e.g. editorial control versus scalability). We also compare the implementation of semantic SEO on top–tier sites such as Buy.com and CNET, as well as up–and–coming sites like WidgetBox or HubPages.
Semantic Web applications often involve integration of information using different ontologies.
The Semantic Web Rule Language (SWRL), a W3C Member Submission, is an effective representation for translating data between different ontologies. This includes mapping of classes, properties, and values, structural transformations, and unit conversions and other calculations.
This talk will introduce SWRL, discuss the open source Snoggle editor and other tools, and demonstrate their use in translating data from multiple independent sources. It will also discuss alternative mapping representations including SPARQL and RIF.
RESTful services are a lightweight implementation of the Service Oriented Architecture (SOA). They wrap resources and talk XML, and have found even greater success than their heavyweight WSDL and SOAP-based siblings. Mashups, which are web applications created by composing two or more services and overlaying the discrete data they provide, have become a popular approach towards creating a customizable Web.
Unfortunately, mashups do not embrace the idea of a read/write Web. The complexity of programming languages such as Javascript, the lack of abstraction of security, and access control principles are a few of the barriers that make it difficult for people to create their own applications. Further, the limited customization capabilities often lead to one-size-fits-all applications. Current tools from IBM, Yahoo!, Google, and Microsoft that help create mashups with reduced programming are limited in the number of services with which they can interact. They normally deal with services that are internal to the company where the tool was developed (Google Mashup Editor, for example, can use Google Maps) or to services that have standard types of outputs, such as RSS or ATOM. Perhaps the most challenging aspect of creating mashups is the ability to stitch together services when data mismatch and heterogeneity are present. Anyone with experience in data interoperability and integration also knows that it is hard to integrate data with purely syntactic and structural means; most often semantic approaches and techniques are needed.
How, then, to address these limitations and find a less complex, yet scalable approach? Reuse and data mediation were the very problems that led to a number of proposals for Semantic Web services and, recently, to the W3C recommendation of Semantic Annotation of Web Services Description Language an XML Schemas (SAWSDL) [[link]]. We build the SA-REST description by borrowing the idea of grounding service descriptions to semantic meta-models using model reference annotations from SAWSDL. However, the lack of a formal representation framework for REST services makes it challenging to add the semantic metadata. Using emerging microformat-based approaches such as RDFa and GRDDL, we propose an approach to add semantic annotations via properties of HTML elements.
In this presentation, we will introduce SA-REST and compare it with SAWSDL,describe tools for easier creation of SMashups with limited programming, and use semantics to do automatic data mediation, give live demos of sample Smashups from popular multi-party RESTful services, and discuss the status of current discussions on SA-REST in W3C process.
Enterprise operations and systems use, change, produce, and discard metadata and data information along with other resources. This diverse operational metadata and data can be accessed and then semantically processed to yield an aggregated 'intelligence inventory' of the operations (IIO). This IIO can then be integrated, used for interoperability, and then unified by using a tailorable general management model, which in turn provides holistic and dynamic situational awareness and response for subsequent operations.
This general management model cycles through the following activities:
Current organizational mission and functional operations Operational and analytical metadata and data Intelligence inventory Intelligence unification (as an advanced form of full enterprise architecture) Intelligence-based applications, security controls, simulation, and decision support Intelligence-based holistic, cohesive, coherent, and concurrent strategic planning (e.g., holistic balanced scorecard) Intelligence-based mission and functional operations
This session covers the end-to-end implementation with lessons learned for a Semantic information portal that is in use today for more than 80,000 documents. The key features include narrowed search, personalization of site navigation and search results, and dynamic integration of content. The portal makes use of Open Source components including the Eclipse Help System, Lucene search engine, and RDF APIs as well as Open Standards such as DITA XML and SKOS RDF. The semantic definition consists of a taxonomy with core facet categories, cross-category relationships, and rich textual synonyms. Points of interest include the following:
Defining the taxonomy and classifying content with semantic hypertext Starting with a broad-and-shallow approach to semantics Focusing on collections of documents rather than on individual documents Integrating the classification into search ranking Mapping textual search input to classification subjects Providing a separation of concerns between display and semantics
In 2007 eMonitor presented on how it was learning the Semantic MediaWiki to manage learning and to align learning with organizational goals. We have now been using this platform for more than a year and use has expanded dramatically, sometimes in unexpected directions. In this presentation we will summarize what we have learned from a year of heavy use by a growing user population.
- Initial goals and how they are being met - New uses, expected and unexpected - Evolution of the semantics - Introduction of forms and templates - The role of metrics
Thanks to Google, Yahoo and others like them, web search technology has been so successful that everyone simply enters a few words into the ubiquitous search box for answers to just about anything. However, the search engine's huge success has also made us accept a long, albeit instant, listing of sometimes irrelevant search results. It won't be a surprise that many consumers think that what these search engines currently provide is the best we can ever get. Are we fooled by the relative success of search engines? What is the next big innovation in search technology?
Agenda: 6:30-7:00 pm Registration and networking. SemTech Exhibits Hall open 7:00-8:00 pm Short presentation from panelists (5-10 min each) 8:00-9:00 pm Q&A
With the recent market saturation and increasing competition in the telco industry, Korea Telecom (KT), which is one of the largest companies in Korea, is making an all-out effort to discover new growth momentum for the future. Conveniently named Telco 2.0, the theme for KT's efforts is to transform itself into a customer-centric and highly innovative company focusing on the five new business areas of contents, solution, cultural complex, knowledge consulting, and computing. The importance of the IT architecture (ITA) in realizing this vision cannot be overstated. The ITA for the future must support effective real-world representation, deal with various levels of heterogeneity, and provide intelligent processing over enriched information. We will present KT's semantic technology roadmap and show how we expect semantic technology to address these areas. We will also introduce KT's semantic search engine STARS developed as part of these efforts.
Software Engineering involves the knowledge representation and translation from mental models to executable language elements. This creates semantic relations between the various representations within software artifacts. These artifacts are expressed in variety of languages ranging from natural language for requirements, UML and XML for modeling, Java / .Net for implementation, DDL / SQL for data storage and query and XML for configuration. These semantic relations constitute the implicit knowledge associated with a software system. Software Change impact analysis is well known end technique to assess the scope and consequence of a change during evolution. Traditional approaches to impact analysis have analyzed the code level; dependencies or the traceability relations between lifecycle artifacts. In the current software engineering context this is incomplete and leads to faulty conclusions because the implicit or semantic associations have not been considered. In this presentation, we describe valuable applications of ontologies, knowledge representation languages and reasoning to problems related to impact analysis. We study and solve two specific problems that we identify as ‘semantic scatter’ and ‘semantic references’. These problems are inevitable during the software lifecycle when two elements (individuals) in two different artifacts realize the same knowledge element (scatter) or when one individual makes references to attributes of some other individual (semantic reference). We first describe our knowledge representation language called link definition language (LDL) that is used to explicitly represent semantic links across software artifacts. Link statements embed concept selectors or rule expressions using existing languages like XPath, SQL and our own logic based formalism called PredQL to associate ontological concepts with language elements found in software artifacts. The link statements are then evaluated against a semantic repository within a relational database to infer explicit relations between individuals that otherwise remain hidden. The application of semantic technology to address knowledge engineering issues, specifically with respect to software impact analysis is the key contribution of our paper. We will demo the prototype version of a solution implemented as a Eclipse plug-in and share a case-study to help our audience appreciate the problems and the value of our solution.
Ontologies contain explicit declarative statements of term meanings. But there are many ways to encode meaning in an ontology, some of which are better than others, as you might expect. Modeling choices may be both domain and application dependent. Reusability also plays a role in modeling decisions.
This session focuses on methodology in ontology development as well as best practices in vocabulary management and provides an early view of what's coming in OWL 1.1. We will cover a number of methodology issues and rules of thumb, discuss emerging policies and good practices for managing and publishing RDF vocabularies and OWL ontologies, and demonstrate the use of development tools, online validators, consistency checking, explanation technology, and advanced analysis techniques. Examples will be taken from scientific, financial services, and e-commerce applications. We will walk through an end-to-end application with light-weight, back-end reasoning, to demonstrate the benefits of the technology, and provide links to online and open source resources for further investigation.
In this talk we discuss:
How and why semantic technologies such as RDF, OWL, and SPARQL are used in the application Using SQL Server as an RDF repository and inference engine via integration with RDF Gateway Semantic Web application development with .NET tools and technologies (e.g. SilverLight and LINQ)
The consequences of man-made (terrorist and ballistic missile attacks) and natural disasters (hurricanes, earthquakes, and tsunamis) can be managed more effectively if tools that enable access to and analysis of complex real-time and a priori data are available to governing authorities. Furthermore, proactive consequence planning before a disaster can save additional lives and property. Cloudwall, a reasoning framework over real-time, federated and service-enabled data, is being developed for the DoD to provide an innovative approach to the problems of consequence planning and management. Issues discussed will include:
* The Cloudwall system architecture
* Domain ontologies and domain reasoning
* Consequence planning
* Consequence management
* Visualization of consequences (with semantic annotations)
This presentation reports on activities over the last two years that led us to the intersection of three exciting areas of IT. Social Software, Enterprise Architecture, and Semantic Technology. Using MediaWiki and the Semantic MediaWiki add on, we have created a corporate wiki we call 'Excellupedia.'
Our original intention was to find a new way to document our enterprise architecture. However, our small project took on a life of its own as we discovered more and more groups in the company asking to be a part of our new experiment. We also found we could use 'the wiki' to capture semantic relationships with relative ease. We will show you examples from our project and demonstrate how the Semantic MediaWiki works.
We will talk about our lessons learned and the best practices we enjoy. If you are considering building a semantic wiki in your organization, then this presentation might be for you.
The Semantic Web is nearing the point of widespread practical adoption:
The core specifications have stabilized Tools and frameworks implementing key features have been through several development cycles An increasing number of major software companies have developed semantically enabled products or are actively researching the space
As companies start to translate theory into real applications, they are confronted with a host of practical software engineering issues:
What is the standard or recommended functional architecture of a semantic application? How does that architecture relate to the Semantic Web standards? Which of those standards are stable and which can be expected to evolve in ways that would significantly impact prior applications? What types of tools/frameworks exist that can be leveraged to help implement semantic applications? How mature are the various categories of Semantic Web tools/frameworks? Can API standardization be expected for certain tool/framework categories? What best practices exist for the design, implementation and deployment of semantic applications? What future trends in support for semantic application development can be expected?
This panel session gathers together semantics experts from the software industry to address these and other practical issues relating to the development of semantic applications.
One big hurdle of building the Semantic Web is to turn unstructured information on the Web to meaningful representations in a semantic space. Semantic Signatures® provide a way to represent free text in a semantic space, modeled on a set of predefined semantic dimensions. Various applications can then be built on top of this model.
1. About Semantic Signatures (10 minutes) (Speaker: Wen Ruan)
Defining the semantic space Building a vector model of the Web Comparing Semantic Signatures to other semantic technologies Building semantic dictionaries for verticals
2. Using Semantic Signatures as tags for a bookmarking application; mapping blogs into semantic space (10 minutes) (Speaker: Drew Farris)
Navigating the blogosphere by category and concept Finding conceptually related matches. Providing concept-level tracking. Dealing with user-defined queries.
3. Using Semantic Signature to generate rel-tags to map to the existing tag world. (10 minutes) (Speaker: Todd Chronis)
Create a semantic signature to identify the subject matter of the document. Identify the highly weighed terms in the document determined by the signature. Filter these terms through existing set of tags for uniformity. Output the terms, with links, to an existing tag set (such as del.icio.us).
4. Using Semantic Signatures to match web pages to ads. (10 minutes) (Speaker: Mary McKenna)
TSVs efficiently generate semantic representations for both web pages and advertisements. Both pages and ads are tagged with dimensions and associated weights, composing unique semantic signatures. Matching is done by comparing signatures using a vector space model. Comparing a web page signature to ad signatures results in relevant ad placement.
5. Using a custom semantic dictionary for eBay search and discovery. (10 minutes) (Speaker: Todd Chronis)
Creating a semantic dictionary to capture auction space. Guiding the user’s query using the semantic dictionary. Navigating auctions by direct signature manipulation. Optimizing result-order.
Part I: Creative Commons has a long history of embracing Semantic Web technologies. From the beginning, our licenses have been provided for three audiences: humans, machines, and lawyers. Licenses are described using RDF, and users have been encouraged to embed machine-readable metadata in licensed pages. Until recently, however, the software which served this information was using a series of XML files and scripts to generate the machine-readable information.
Since summer 2007, the software stack has been evolving to the point where we 'eat our own dog food': new software such as our OpenOffice.org plugin are driven using the same RDF and machine-readable information we serve to our users; the license engine is evolving to use the same dataset.
This presentation will provide: An overview of our infrastructure stack, then and now Specific examples of how we've replaced pieces of our stack with semantic counterparts An overview of the ways in which using semantic technologies have saved us time and energy An overview of using tools such as the Semantic MediaWiki to begin leveraging the Semantic Web immediately
Part II: In 2002, Creative Commons introduced copyright licenses to help foster collaboration and reuse. Our work has always pushed three parallel efforts: the drafting of legal text in various jurisdictions and for various purposes, the development of comprehensive and easy-to-understand resources for non-lawyers, and the development of technical means to express licensing information in interoperable, machine-readable ways, so that reuse and collaboration could be made even more efficient. On this last front, machine readability, Creative Commons chose the W3C's Resource Description Framework, but was left with only poor technical means to successfully include RDF in HTML pages, the predominant Creative Commons medium.
After a long collaboration with the W3C and other partners, Creative Commons is introducing ccREL, the Creative Commons Rights Expression Language, a comprehensive technical recommendation for embedding RDF in Web pages and standalone files. ccREL is built to be easy to publish, easy to parse, and easy to transfer from one work to the next at the point of reuse. Thanks to the properties of RDF, ccREL is also designed to be truly extensible at the semantic level, so that Creative Commons licenses can be applied to entirely new categories of works we do not yet envision, without changing the basic architecture or breaking existing tools.
In this presentation, we will cover: The design principles behind ccREL Why Creative Commons first selected, and is now continuing to use, the W3C's RDF Creative Commons's work with the W3C to produce RDFa, a mechanism for embedding RDF in HTML that reuses rendered content for semantic purposes Examples of ccREL uses, and how existing services could benefit from implementing ccREL Guidance for developers on finding, parsing, and publishing ccREL
Thomson Reuters evangelist Thomas Tague leads a focused discussion of the Calais Web service and the thriving Semantic developer community at OpenCalais.com (www.OpenCalais.com) The Calais Web Service enables publishers, bloggers and sites of all kinds to automatically metatag the people, places, facts and events in their content to increase its search relevance and accessibility on the Web. It also lets content consumers, such as search engines, news portals, bookmarking services and RSS readers, submit content for automatic semantic metatagging that is performed in well under a second. Join Tom to learn five easy ways to kick-start Semantic metatagging for your site or service. Gain valuable insight on effective ways to create customized Calais tools and applications to ensure your content is more easily integrated into social networks, widgets and semantic applications of all kinds.
Part I: Analyzing Web Access Control Policies Using Semantic Technologies Policy management is becoming an increasingly important IT task in large, complex organizations. With the broad move away from imperative systems and programming languages, toward declarative systems and policy and configuration language, some IT problems are solved, but new ones are inevitably introduced. Policy-based systems have lots of advantages over programming language-based systems; but policies have to be managed, like every other information technology artifact.
The problem arises when the 'semantics' of policy languages are either complex or not well understood or both. In those cases, in order to manage the policies themselves, IT departments are typically faced with the expensive, burdensome task of building policy management systems from scratch, which takes their focus away from their primary tasks, without any confidence that these management systems are actually 'correct.'
Our system focuses on policy management in three crucial areas:
Policy development Policy testing Policy audit, verification, and forensics
In this demo, we will show policy management in the area of web services, specifically of XACML and WS-Policy policies. The analysis services that our tool offers include:
Formal policy verification and deep testing Policy change analysis Policy redundancy Policy repair, debugging, explanation Policy federation (for example, disjointness checking) Policy set optimization
Part II: Shared, Ontology-Based Access Control for an SOA Traditionally, enterprise access control has been implemented on an application-by-application basis. The resulting security infrastructure is often redundant, complex, and difficult to change. Recently, SOA advocates have been pushing the idea of 'shared security' so that access control policies can be consolidated into a single, shared service.
This session presents an implementation of that idea that uses OWL ontologies to represent the actors, resources, and security policies of an enterprise. When a request is made, these ontologies are combined and submitted to an OWL reasoner for analysis. Based on the information it receives, the access control service can either approve or deny the request. However, it also makes use of the open-world assumption in OWL. If the service decides that it does not have enough data to render a decision, it can request more information. This allows the security service to work with an incomplete representation of the system it is securing, bringing in more data on an as-needed basis.
This Opening Exhibit Presentation will explain how the world’s largest enterprise software company is bringing semantics into the mainstream. Databases, Middleware, Governance Applications and ERP Systems will all soon be delivered with semantic technology under the hood — find out why. Additionally, this presentation will articulate the W3C’s guidance for developing your Business Case for semantic technology and offer a preview of explaining the Semantic Web for Dummies.
Representing uncertain and incomplete knowledge in ontologies has a wide range of applications. However, none of the languages in the Semantic Web stack--i.e. RDF, RDF-S, OWL--provide the means to express such knowledge. In this presentation, we will describe how probabilistic knowledge can be represented in OWL ontologies and demonstrate the probabilistic reasoner Pronto that extends the OWL-DL reasoner Pellet. Pronto can reason with ontologies where a probability interval can be assigned to statements, specifying the probability that statement is true.
Our presentation will focus on how probabilistic ontologies can be used to solve real-world problems. We will provide examples and use cases from different domains and application areas with a specific focus on the following three cases:
Probabilistic ontologies for risk assessment in healthcare sciences Representation of probabilities in social networks for intelligence analysis Data integration from heterogeneous sources by utilizing probabilistic ontology mapping
NeOn addresses the complete R&D cycle of the emerging new generation of semantically enriched applications, which exist and operate in an open environment of highly contextualized, evolving, and networked ontologies. NeOn aims to achieve and facilitate the move from feasibility in principle to concrete cost effective solutions, which can support the design, development, and maintenance of large-scale, semantic-based applications. In particular, the project investigates methods and tools for managing the evolution of networked ontologies, for supporting the collaborative development of ontologies, and for the contextual adaptation of semantic resources.
NeOn is creating an open, service-centered reference architecture for managing the complete lifecycle of networked ontologies and metadata. This architecture is realised through the NeOn Toolkit and complemented by the NeOn methodology for system development using networked ontologies.
NeOn pays special attention to integrating research into work practices. Therefore, the methodology, toolkit, and infrastructure are intertwined with their deployment and testing from the early phases of the project. NeOn uses a case-centered methodology, which means that our research results are applied to real-world case studies involving partners from industry and public bodies. One such case study is using NeOn technology to develop an Ontology-driven stock over-fishing assessment system at the Food and Agriculture Organization of the UN, which focuses on the agricultural sector and information management for hunger prevention.
In the talk we will:
Describe application scenarios and use cases for semantic technologies Discuss emerging trends in next-generation semantic applications Present NeOn technologies for engineering ontologies and developing semantic applications, covering the complete ontology and application lifecycle Demonstrate the use of NeOn technologies in a concrete case study, developing an Ontology-driven stock over-fishing assessment system
All mobile service companies have invested big money in data service infrastructure, but only a few people are using data services with their mobile phones. Therefore, telecommunications and mobile service companies are looking for a new 3G-based business model to increase their revenue by expanding service buying power through personalization, context awareness, and service adaptation.
In this talk, Saltlux will present a new semantic-based mobile service model and propose a middleware system based on IMS(IP Multimedia Subsystem) architecture. Saltlux has developed new, innovative middleware system for intelligent content delivery enabling (CDE) by applying ontologies and reasoning with rules. This presentation will include the real use-cases and service scenarios of KTF. The biggest Korean telecom company, KTF, introduced the CDE system for their future service infrastructure.
Video is becoming increasingly important in the Web. Analysts predict that by 2010, 85% of the traffic through Internet routers will be video. This will create demand for a new breed of data management software focused on video, not just structured record data, and the capability to search and stream video content.
Where does the Semantic Web come into this? Film studios will continue to try license viewing and purchase of their films. But TV studios, already used to ad-supported networks, will probably move quickly to ad-supported distribution models on the Internet. This means that ads will have to be placed in or near video. But the process of determining what is a contextually relevant ad placement in video is harder than keyword-based placement in textual Web pages. Video will require a semantic markup that not only identifies product placements in the video, but more deeply, says what is going on in the video, what actors are wearing, riding, drinking, doing, feeling -- Semantic markup of video will become the key to ad-placement in video. This talk will discuss a new DBMS product that supports both the high-level ontology-based semantics of the W3C's OWL, but has also been designed with a class-specific storage system that allows it to handle both structured metadata and the actual video content and includes timecode-based indices that allow it to pinpoint the location of specific scenes, shots, actors, and locations within a video. We think that these new video databases may find a totally new market almost as large as the existing DBMS market for structured data. They will not return tuples to Java programs running in an App Server. They will stream data to thousands of viewers running browsers or Internet TV across the Web, to iPods, and video-enabled phones or bundle it for delivery to edge-server networks along with embedded hotspot advertising. This is a brave new world for data management: Optimizing SQL is just no longer interesting.
Smart Browsing is about wiring recognition of everyday things into the browser. Once the browser detects that user interacts with books, music, movies, people, recipes, etc., it can be smart and helpful about connecting user to related information. Learn how AdaptiveBlue technologies leverage a mix of standards and common sense to enhance everyday browsing.
Many enterprise databases contain information about people, companies, relationships between people and companies, places and events. The Semantic Web carries the promise that we can use Semantic Technologies to analyze networks of people and companies and events in time and space.
In our presentation, we will show how we can combine the current state of the art for RDF and OWL reasoning with Allen's Temporal Interval Logic, some basic Geospatial primitives, and a number of well-known Social Network Analytics. This is a new and interesting combination that allows for new types of Business Intelligence.
In this presentation, we will discuss a simple event ontology that enables geo/temporal/social reasoning. Some common elements that we see in all event databases are that an event always has a type and possibly some subtypes, one or more actors, a start time and possibly an end time, a location specified symbolically as a place name or address or using longitudes and latitudes.
The presentation will demonstrate new query capabilities using several real life examples.
What should you know before you set out to build a robust ontology for your business? And when are your existing controlled vocabularies good enough? Join Wendi Pohs and Breanna Anderson as they share their experiences building enterprise ontologies in large corporations. We will discuss both business and tool requirements for creating an integrated, semantic infrastructure. We will survey existing tools, and we will share some of the more common implementation techniques and pratfalls, including our own.
Travelers are demanding more information from trustworthy sources (other travelers). That information is easier to find and use if we begin to describe our reviews, trips, and plans in a common language. Microformats provide this necessary glue and can be adapted for describing trips planned or taken. Tripblox is a prototype web application that accepts 'pings' or notifications indicating URLS where semantic data exists. It then consumes the feed and provides the data back to travelers. In the process, it collects an inventory of activities, lodging, and points of interest.
Attendees will be given ideas and solutions for the following:
How to include geonames data in Semantic Web applications Representing 'nearness' within OWL ontologies for findability Parsing microformats from xhtml Binding Java to RDF via jenabean, an open source library Using simple reasoner techniques to add new data upon what has already been asserted
Semantic Web standards and technologies already gained popularity as a platform for data integration and analysis in the life science (LS) domain. The size and diversity of the LS databases makes them one of the most challenging areas of application of semantic technologies: there exist over 800 public biological databases; the RDF representation of UNIPROT contains over 1 billion statements.
LARKC is a research project aiming at development of 'web scale and style' reasoning methodologies and tools. The Protein Interaction Knowledge Base (PIKB) will be developed within the project by means of syndication of several of the most commonly used databases, e.g. UNIPROT, Entrez-Gene, GO, BioGRID. PIKB will serve as a show-case for the LARKC technology — a demonstration server will allow the general public to query to explore PIKB. The presentation is organized as follows: an overview on scalable reasoning approaches an early version of PIKB, containing about 2 billion statements demonstration of OWL reasoning and query evaluation on top of PIKB and discussion on the methodology for adaptive data syndication using formal semantics Discussion on the usage of PIKB is used for semantic annotation of life science articles demonstration of hybrid semantic queries and a faceted search interface
Your future knowledge store will be a collection of inter-related triple-stores federated into an ever-changing whole. Scaling your application in Amazon's EC2 is now easier than ever with AllegroGraph's flexible architecture for creating 'federated knowledge stores' on the fly. This EC2 enabled architecture allows for transparent RDFS++ reasoning and SPARQL queries across Billions of RDF triples. With AllegroGraph's GeoTemporal and Social Analytics package it becomes the core platform for scaling your semantic web application.
Learn how you can leverage the power of EC2 and AllegroGraph by attending this presentation.
In today’s world of "information overload," even simple information analysis can become overwhelming. Since data typically comes in a variety of text forms, the information may be fuzzy, incomplete, and all too-often is a literal pile of paper to read. Furthermore, business users often spend more time cleaning, sorting, and reformatting data in preparation for analysis than analyzing the data itself. The Metatomix Analytic Platform provides a wealth of tools for attacking this information overload. By allowing users to semantically cluster, categorize, and analyze huge volumes and varieties of data sources, the Analytic Platform increases the amount of information a person can process and review at one time.
Components and capabilities of the Analytic Platform:
Metatomix Analytic Studio – workgroup-enabled document organization & analytic framework Metatomix Text Analytics & Visualization – unstructured text visualization & analysis Metatomix Parsing Agent – document normalization and transformation tool Metatomix Automated Document Categorization – user-driven document clustering & visualization Metatomix Semantic Document Analysis – semantic extraction, visualization & enhancement with support for enterprise RDF stores
See description on attendee message board.
(This session continues from 5:30PM - 7:15PM) This one day event is an exciting development for us at TopQuadrant. TopBraid Composer, the first TopBraid Suite product, available since May 2006 now has a large, growing base of users from hundreds of enterprises and organizations spanning countries across the globe. It is considered by many to be the leading professional ontology modeling and development tool. During the past year we introduced several new TopBraid products (TopBraid Live, TopBraid Ensemble, and the Maestro Edition of TopBraid Composer), significantly expanding the capabilities of TopBraid Suite. All components of the TopBraid Suite continue to work within an evolving, best of breed, open architecture platform built specifically to implement W3C semantic web standards. Along with the expansion of the TopBraid platform, we have also seen a considerable growth of the user community and strong interest in sharing experience and best practices for building semantic-enabled applications.
The day’s activities and agenda will feature:
· Presentations and demonstrations from the TopBraid user community sharing their experiences in developing ontologies and building semantic web applications
· Presentations by the TopQuadrant team on new product features and best practices
· Opportunities for Q&A and networking
We are pleased to be able to conduct this event in conjunction with SemTech2008 — a leading venue for the sharing of experience and insights between participants in the emerging semantic web marketplace — as we believe it will allow more TopBraid users to attend and benefit from mutual exchange of ideas, questions, effective patterns of use, knowledge, views and suggestions for future directions and enhancements.
- 10:45 – 11:00 Welcome
- 11:00 – 11:45 Keynote User Talks
(two 20 minute presentations by selected TopBraid users)
- 11:45 – 12:45 Overview of Product Features and Best Practices:
"TopBraid Suite 2.0, Capabilities to Support Information-Integration-Intelligence",
Holger Knublauch, VP of Product Development, TopQuadrant
- 12:30 – 1:30 Lunch
- 12:45 – 1:30 User Feedback Session (conducted through lunch)
- 5:30 – 6:05 Demonstrations of Use of Key Capabilities
(two 15 minute selected demonstrations of use by TopBraid users)
- 6:05 – 7:05 Panel Discussion, Semantic Tech Solutions: Strategic Capabilities and Tool/Platform Requirements
(Panel will consist of members of TopQuadrant and Selected Users)
- 7:05 – 7:15 Announcements and Closing Remarks
This presentation highlights linguistic approaches to sensitive data detection and protection.
This report presents rationales and results of a working group with members from industry and academia in translational research. The Working Group’s goals are to develop strategies and toolsets using semantic approaches to better understand multiple data from —OMICS, tissue imaging and clinical trial results as well as public resources within their biological relevancy.
The Relational model of data is well understood. We know how to deal with such data, and years of practice have taught us how to design and build software applications based on relational databases. Nowadays generic tools, libraries, and frameworks allow us to rapidly develop new customized solutions.
However, semantic data has a different nature. The structure is more flexible (a graph), and information is expressed in the form of statements (triples). Furthermore, we do not necessarily have fixed schema, and thus the data can be more unpredictable.
I'd like to present how to deal with such information, how to effectively process it, and how the design of our software applications should change to fully benefit from a semantic data model.
This presentation consists of the following steps:
Current practice and model of dealing with relational data in software applications Problems and limitations imposed by the model when building semantic data based applications Observations, remarks, and suggestions for design improvement Solution: demonstration of SemanticObjects - a library mapping semantic data into interconnected Perl objects automatically Example: the design of IDEAS (Individualized Digitized Educational Advisory System) - a semantic data based application using the SemanticObjects library
Semantic Web data is modeled as an arbitrary graph of interconnected entities which is constantly changing and growing in size. The type of entities and relationships between them cannot be known a priori. An important characteristic of the semantic graph is the fact that some of the relationships between various entities are not explicit (they are inferred) and can appear between any parts (sub-graphs) of the semantic space. Designing a user-friendly and comprehensible graphical user interface for such cases is not a trivial task using traditional approaches. The Atom Interface, which we have developed, allows users to navigate the semantic space in a user-friendly manner; the user can focus on navigating different parts (sub-graphs) of the data set at once and see the connections between the different parts whether they are explicit relations or inferred ones. We will present the following:
The model of Semantic Web data State of the art in browsing the semantic space (traditional and non-traditional approaches) Challenges in visualizing and browsing interlinked graph data Presentation and demonstration of the Atom Interface Conclusions
Ontology is a fundamental data object for organizing knowledge in a structured way in many areas ranging from Philosophy to Knowledge Management. Knowledge capture, knowledge integration and knowledge delivery are the essential parts of dynamic knowledge management systems. In this talk, we present a brief description on the role of ontology in knowledge management and review the ontology building tools for Information Architects (IAs). The purpose for reviewing ontology building tools is to determine the toolkit most suitable for ontology creation, editing, and mind/concept mapping from the view points of Information Architects (IAs) who play a significant role in designing knowledge management systems. This presentation also gives a fundamental understanding of ontology tools available on the market as open source products as well as commercial products in terms of their capability, availability, enhancement and further development. We provide a ranked list of the tools based on our needs and suitability for the IAs. The presentation will include discussions on the following key subtopics.
* What is Ontology and why do we need it * What is Mind Map and its usefulness * The role of Information Architects and Knowledge Architects * The relationship between MindMap and Ontology * Some examples of Mind Mapping Tools * Why MindMapping tools are not enough to describe an ontology * A review of ontology building tools * A ranked list of Ontology building tools based on specific criteria * Conclusions and recommendations
Since the initial standardization, Web Ontology Language (OWL) has acquired a rich infrastructure of tools and a wealth of experience in areas as diverse as medicine, biology, geography, astronomy, defense, and the automotive and aerospace industries. However, it is not always obvious how or when to apply OWL. People who are not familiar with OWL have difficulty understanding how OWL reasoning works and/or can be used in practice. The aim of this tutorial is to help participants make informed decisions about whether OWL and reasoning has anything to offer their application. We will specifically focus on the OWL-DL reasoner Pellet that comes with features such as consistency checking, query answering, explanation and debugging, rules support, and incremental reasoning. We will describe these features in the context of real-world problems that range from ontology development in health sciences to access control policy management and development.
Semantic Web technologies are often perceived as difficult by the typical developer, despite the simplicity of the conceptual models. In reality they are difficult, not least because the technical details of formats and protocols obscure fundamental patterns. It will be argued that when viewed from different perspectives the Web itself can provide abstractions to simplify development.
A central part of the HP Software portfolio is the HP Universal Configuration Management Database. The UCMDB is an enterprise scale metadata repository that supports EII by providing a single point of access to dynamic models of business services and their supporting assets. This talk describes recent work on adding a SPARQL interface to the UCMDB, enabling lightweight programmatic access via the web. In addition we talk about efforts to harmonise the ontology used to describe IT Service Management and how this is key to enabling business users to query the UCMDB in their terms.
The interfaces we use to interact with the world's information are getting smarter. Web portals gave us someone else's idea of the content we should see. Then came search engines, which let us tell the system what we want, one query at a time. We are about to see the next wave -- intelligence at the interface -- in which the system knows about us, our information, and our physical environment. With knowledge about our context, an intelligent system can make recommendations and act on our behalf.
The Net-Centric Operations Industry Consortium (NCOIC) comprises a who's who of defense, intelligence, and aerospace contractors. It's mission is to collaboratively engineer new concepts and approaches that will enable the US, its allies, and industry to transition from an era of essentially stand-alone systems and data silos into a world where organizations, personnel, systems, information, and other assets dynamically interface, interoperate, and communicate in context across the Global Information Grid (GIG). A very tall order. This session features three perspectives on the challenges, progress to date, lessons learned, and the role of semantics in delivering net-centric interoperability.
Steve Russell will provide an introduction to NCOIC, emphasizing the social aspect of net-centricity and it's role in establishing 'shared situational awareness' across the community regarding what it means to be net-centric, what the full spectrum and complexity/diversity of the problem space really is, and how this impacts the NCOIC approach to interoperability.
Hans Polzer will discuss the Systems, Capabilities, Operations, Programs, and Enterprises (SCOPE) model for interoperability assessment as a way to describe/measure the diversity of the net-centric ecosystem, both from an institutional scope/perspective viewpoint, and from a degree of system/network open-ness viewpoint.
John Yanosy will overview the semantic approach being taken towards pattern development and characterization, why semantics are needed, and the work that has been done to date.
Imagine if video game programmers had to write (or, worse still, reinvent) texture mapping, pixel shading, 3D rendering, or floating-point math code for each game they developed. Either the cost of game development would increase a thousand-fold, or we'd all still be playing Pac-Man and Space Invaders. While graphic-, math-, and physics-coprocessors have dramatically improved the state of video game (and many other) applications, the lack of a comparable semantic coprocessor has severely constrained improvements in the intelligence of software systems. To date, three factors have hindered the development of such easily reusable semantic processing capabilities: Lack of standards for representing and exchanging semantic information A less-than-critical mass of information in machine usable form Insufficiently powerful and efficient reasoners to support the effective use of a large corpus of semantically rich information.
Recent work in semantic technologies has made significant inroads in all three areas: the community is converging on a handful of knowledge representation standards (OWL, RDF, SPARQL, etc.) and information interchange architectures (SOA, web services, etc.); numerous ontologies have been developed (ranging from insular domain-specific ones, to federated ontologies, to more broad-based and comprehensive ones); and advances in efficient machine reasoning, combining traditional statistical and semantic approaches, are showing promise. This talk describes how software can leverage these advances and presents examples of applications that behave more intelligently as a result.
In many Semantic Web applications e.g. health informatics,security informatics, e-commerce, e-government, social networks, and community information systems, the need for sharing information among autonomous entities needs to be balanced against copyright, privacy, security, or commercial concerns. This calls for mechanisms that enable the participants to selectively share knowledge with others without risking the disclosure of hidden knowledge, i.e. knowledge that needs to be protected from disclosure.
We argue that current policy languages for information hiding on the Semantic Web that prohibit access to hidden knowledge are overly restrictive in that they prohibit the use of hidden knowledge in answering queries even in scenarios where it is possible to do so without disclosing hidden knowledge. We describe an approach to selective sharing of ontologies, in particular, answering queries against an ontology based on inferences that use hidden knowledge, without compromising hidden knowledge.
This talk, aimed at a broad audience of semantic technology practitioners, will:
Motivate and introduce the problem of privacy preserving inference Describe a general framework for privacy-preserving inference that exploits the indistinguishability of hidden knowledge from incomplete knowledge on the Web Describe and illustrate, through examples, a set of provably privacy-preserving inference strategies for a broad class of Web ontologies, including widely used hierarchical ontologies Show how privacy-preserving inference engines can be implemented as extensions to existing reasoners for commonly used ontology languages
The Semantic MediaWiki was designed, in part, to author and extract (semi-)formalized ontology from a 'traditional' wiki editing environment. We have discovered that this environment is also well suited as a platform for the publication, review, and extension of existing ontological content. We describe how we have extended the Semantic MediaWiki with a standardized terminology model and a service-based back end which allows us to publish multiple inter-linked ontologies, accept corrections and extensions, and transfer these suggestions to a Protege-based ontology editing environment, and report on the successes, pitfalls, and lessons learned to date.
This provocative session asks, 'Just how much can we expect to achieve by embedding semantics into an enterprise's information architecture?' Many vendors (and, we fear, a fair number of their customers) are touting semantics as 'the solution' to a bewildering variety of problems. Over the past 2 years the U.S. Air Force--certainly a complex organization with heterogeneous content and users ranging from novice to expert--has undertaken implementing a semantic service-oriented architecture (Semantic SOA) to meet its future information retrieval and discovery needs. In this session, we provide an overview of the Air Force's expectations and methodology but concentrate on lessons learned: How well is the Air Force meeting its own requirements? How much weight can semantics bear?
Governance approach Ontology alignment Benefits of semantics Limits of semantics Cost vs benefits: 'Would we implement semantics again?'
Many Semantic Web technologies rely on web page authors to provide content in RDF, so they can become part of the Semantic Web. This adoption process is slow and is hindering the growth of the Semantic Web. To jumpstart the Semantic Web, existing websites need to be transformed into semantic websites.
This presentation provides techniques for turning metadata-rich websites into semantic websites. The presentation covers the following topics:
Creating simple ontologies from website metadata Prepping web pages to enable GRDDL processing Using GRDDL to harvest RDF from web pages Showing how harvested data can be queried using SPARQL
Case studies involving the Internet Movie Database, the CIA World Factbook website, and Edmunds.com are investigated.
The 'intelligence pipeline' - the flow from raw data to managed intelligence - should be continuous, rapid, reliable, responsive, and highly-automated. It's none of these, largely due to failure to share and leverage semantics. We attack the root cause by collaborative OWL-based vocabulary management, but we need tools, traction, and leverage to surmount the obstructions that progress exposes. While vendors can't be expected to overcome cultural, organizational, and process barriers, we do expect more focus on gaps in technology, tools, and understanding. We're disappointed by excessive focus on single-point solutions that result in more silos and too little work on enabling technologies for enterprise-level exploitation of data and intelligence assets. Emerging ontologies must be harnessed to direct and harvest results of disparate extraction and analysis activities - to align, integrate, and resolve entities to real world referents. Semantically-enabled Master Data Management, data governance, metadata management, data-as-a-service abstraction capabilities, and other enterprise information management approaches are needed to resolve disparate representations and leverage semantics across operations. We issue a challenge to vendors to get out of the sandbox help us tackle real-world problems.
In this session, we will describe the experiences and lessons learned in creating a pragmatic platform for developing Semantic Web applications. With the goal of making consumer-oriented software, we will describe our decisions on applying semantics in a way as to provide the greatest benefit to the user and the least impedance mismatch for developers. As most developers today are familiar with object-oriented development, it was important to provide an object-oriented view of the world while still making the underlying semantics accessible. We will give an overview of our technology stack, from data storage to a web development environment custom tailored for semantic development. Developers will also be given an introduction to the published API's for interacting with Twine and our platform.
This panel focuses on the transition of commercial off the shelf software to semantic technologies. Key questions include:
What are the trends and directions towards semantic-enabled enterprise COTS products and services? Why will this happen? What are the drivers for COTS applications to get semantically enabled?
When and how rapidly will this happen?
What does this mean for enterprise customers? What does this mean for IT vendors?
What are the investment plays here, if any?
A starting point is the middleware stack where SOA is becoming SSOA in order to fulfill the promises — We see this in Oracle, Microsoft, IBM, Software AG, and other companies that we know or are working with. The objective is for attendees to understand how this transition will develop, accelerate, and unfold, and what they can expect from major IT vendors over the next several years.
Tasked with consolidating numerous Web publishing systems and data models into a single enterprise architecture, The Content Management Engineering (CME) department at Sun Microsystems is responsible for the core of the company's global Web presence. CME publishes a variety of information from a diversity of sources, so in order to streamline workflow for complex business processes, and to maximize reuse of content, the department has developed a library of source content XML, governed by an XML schema portfolio. This schema is driven by an underlying data model, bound to content elements through annotations in the schema, and tied together by the SwoRDFish product metadata system.
The data modeling combines elements of classic enterprise data architecture with developments from the age of Web and agile development. It has to connect XML schemata to other systems ranging from ERP and traditional data warehouses to stand-alone relational databases to N-Tier enterprise Java systems to flat file collections. It also has to control identifiers for products and the key methods for product tagging and classification, which is driven by SwoRDFish, one of the first RDF metadata systems ever put into production. The problem space is complicated by business factors such as globalization, e-commerce features, and large changes to many of the systems that feed CME.
This opens up many problems for data architects. Some of these (versioning, governance, etc.) are well discussed in the industry, and some of them required CME to chart its own suitable practices, and often to look for delicate compromises. One example of such problems is how to deal with the modeling of content fields that have no natural length limitations, but which may be constrained by implementation limitations in legacy systems.
Semantic technology cannot exist in a vacuum and has to work into existing content and workflow systems. This presentation outlines semantic modeling techniques that have proven successful as the scope of the web venues and responsibilities of CME have steadily increased. It discusses the more interesting problems for data modelers and XML professionals raised in CME's projects.
The Financial Sidebar at the Semantic Technology Conference is a panel and forum geared to:
Provide insight and discussion of niche tools and applications for financial services related purposes; Investigate possible areas of cross-utilization and application, collaboration, and cooperation between customers, developers and researchers; Enable networking opportunities for financial services oriented members of the community who are interested in semantic technologies.
Following the scheduled Dow'>[link] Jones, JP'>[link] Morgan and Guardian'>[link] Life presentations on the morning of Tuesday, May 20, the Financial Track will continue into the afternoon through this panel and forum. Developers and customers will briefly present their semantic implementations/solutions as applied to the financial space, followed by facilitated Q&A.
The perspectives of these presenters are expected to be quite varied, from EDI and transaction-related processes within the sector, to information collection/harvesting, classification, and storage – while others are front-end oriented, in terms of access and usage of information, custom content generation and publishing. The participants are expected to come from a range of perspectives, including banking, insurance, publishing, asset management and related consulting services.
For the presentation portion, to ensure the addressing of a wide range of ideas and solutions, each presenter will be restricted to 10 minutes to provide an overview of their particular context and describe the application of their solution.
Note: All Sidebar participants must be registered attendees of the conference. The Sidebar panel/forum is distinct from the conference’s Semantic Solutions sessions, in that the Sidebar will be more intimate, informal and conversational. This is not a "regular" presentation, but rather a sharing of ideas and perspectives around focused application possibilities – with interactive participation to investigate those possibilities.
Through examples, this session will examine how, and in what process categories, semantic capabilities are being leveraged within the Financial Industry. We will then discuss where else, within the broad context of the financial space, these approaches may be applied.
2:00-3:15 * Introduction and Ground Rules
* What are People Doing Semantically in the Financial Space Right Now?
Each presenter provides a brief (10 minute) overview of their perspective (where are they in the space, how/why are they leveraging semantic technology, and to what end). (Participants grouped into category to which they’re applying SemTech: Messaging / EDI / Transaction / Execution; Data Harvesting; Storage / Architecture / Authentication; Product Creation / Delivery; Search / Matching / Interpretation; Front-end Usage / RealTime Retrieval, Custom Content/Publishing/News, etc.)
3:15-3:45 * BREAK
3:45-5:00 * Conversation
Led by a Facilitator: Discussion around the variety of ways the participants have demonstrated they are applying semantic technologies within the financial community. What are the issues and opportunities in/for the Financial Sector? How can some of the examples be applied to other areas of interest within finance. Are there opportunities for cooperation? Do people want to cooperate? Are more standards needed? If yes, what are they? What tools are needed? What limitations are there (antitrust, perceived competitive advantage, etc.).
Web services are often complex involving hundreds of processing components, XSL transforms, and artifacts such as XML schemas and WSDL documents. Designing, implementing, and maintaining web services is confusing and time consuming without tools to manage the complexity.
This session describes the web service components needed to support OWL-S and demonstrates a home-grown visual web services editor that helps architects and developers wrestle with all of the implementation details. The session includes:
An overview of the OWL-S enabled components within the web service A demonstration using the visual editor to build processes and groundings A discussion of the visual editor output (WSDL documents) A summary of quality and productivity gains achieved using the visual editor
In order to plan for the future state of an organization, business and technology managers need to aggregate available data from numerous information sources. Here we demonstrate how, by defining a business process taxonomy and mapping it to the application architecture, one can answer or infer the following:
What are the functionality gaps & overlaps between applications? Impact analysis: for instance which applications are affected when a new regulation is introduced? What are the quality of service attributes that need to be supported by applications in order to support a particular business process requirement? Temporal analysis: what is my migration path: current vs strategic view of the business process?
During this session, we will also cover some of the challenges we faced in this process and how we overcame them.
As semantic technology emerges as a driving force in information technology, what is its relationship to service-oriented architecture? At first glance, these approaches appear to focus on such different aspects of enterprise architecture that there is little to either worry or to get excited about. Given finite IT resources, however, the two approaches can be seen as competitors in driving architectural strategy. Alternatively, the synergy between semantic technology and SOA could open transformational opportunities. Based on real-world experience, this presentation focuses on how these technological strategies can be coordinated to maximize opportunity and minimize risk.
Topics to be discussed include:
* What are the real maturity levels of SOA and semantic technology standards, tools, and best practices? * Should semantic ontologies technologies replace, yield to, or co-exist with POX (Plain Old XML) in canonical models? * Can semantic ontologies supersede today's UDDI as a service discovery solution? * How can SOA be effectively combined with semantic technology to be more powerful than either in isolation?
The information explosion in biomedicine makes it difficult for researchers to stay abreast of current biomedical knowledge and to make sense of the massive amounts of online information. Images are proliferating at an explosive pace, similar to other types of data in e-Science. Technological solutions are needed so machines can help people to access and use images optimally. While Semantic Web technologies are showing promise in tackling the information challenges in biomedicine, less attention is focused on leveraging similar technologies in imaging. We are developing methods and tools to enable the transparent discovery and use of large distributed collections of medical images in cyberspace as well as within hospital information systems. Our approach is to make the human and machine descriptions of image pixel content machine-accessible through annotation using ontologies--specifications of the entities, their attributes, and relationships among the entities in a domain of discourse. We have created an ontology of image annotation and markup, specifying the entities and relations necessary to represent the semantics of medical image pixel content. We are creating a toolkit to collect the annotations directly from researchers and physicians as they view the images on medical imaging workstations. Image annotations, represented as instances in the ontology can be serialized to a variety of formats, enabling interoperability among a variety of systems that contain images: medical records systems, image archives in hospitals, and the Semantic Web. The ontology-based annotations will enable images to be related to non-image data having related semantics and relevance. Our ultimate goal is to enable semantic integration of images and all the related scientific data pertaining to their content so that researchers and physicians can have the best understanding of the biological and physiological significance of image content. Ultimately, new Semantic Web will emerge that will help people exploit the wealth of images now at their fingertips.
How does a company with massive amounts of real-time data and high availability requirements migrate to the Semantic Web? Very thoughtfully! Join this session for a 'day in the life' view of content from creation through processing and on to delivery. Along the way, learn how to efficiently cluster processes and technologies to intelligently migrate from legacy systems to the Semantic Web. Learn how taxonomies and ontologies integrate with entity extraction and rules-based processing to enable near real-time delivery of business critical multimedia content from thousands of sources. See how Dow Jones uses Synaptica, the 1.5 million+ proprietary codes of Factiva Intelligent Indexing, as the foundation of a technical stack that allows content to be processed in less than a minute in a 24x7 global coding and publishing operation.