| Accepted Papers |
 |
 |
 |
|
CASE STUDIES I (Semantics in the Enterprise) 1. Business Semantics-driven Integration of Service-oriented Applications | Pieter De Leenheer, Collibra nv/sa Business semantics define the contextual meaning of key business assets for your organization in terms of business facts and rules. Business Semantics have a dual utility: the derived business semantics do not only provide a shared glossary to augment human understanding, but can also be used to automate otherwise costly and error-phrone data integration during complex business process integration.
Portals are not new but are typically based on aggregation of structured information. Highly innovative industries like the one for Nanotechnology need to integrate information from various sources with a large portion being from unstructured text. This session will demonstrate how semantic technologies, especially text mining tools, allow to "semantize" content and to interlink pieces together for better decision making. Different stake holders need different information and the presented "Nanotrack" portal links patents, scientific papers, news, blogs and other sources together. In the session the user will learn how implicit connections are identified, how possible trends are recognized and how the system supports a research team in the identification process of skills of persons and companies. The underlying technology is a framework for crawling, ontology driven information extraction and a scalable RDF store.
We, the people, intelligently manage information in a variety of ways. The birthright of ubiquitous computing, of the Internet and of mobile devices (such as smartphones, handhelds and PDAs) renders the marketing of new products and services possible to be embedded within the knowledge economy.
We employ a scenario describing the issue resolution process in the aerospace engineering domain to determine requirements for enhanced knowledge management and intuitive analysis. We discuss the use of Semantic Web technologies in an interactive knowledge framework to support the stages of the knowledge management lifecycle: knowledge generation, manipulation and use, enrichment, sharing and re-use. The framework makes use of a modular domain ontology that defines users and their tasks, and the physical objects users interact with during their normal working activities, to guide the hierarchical, context-driven hypothetical analysis performed during issue resolution. The ontology also guides the enrichment of the knowledge generated as a by-product of users' interaction with the knowledge framework. An important requirement was tool design that improves users' analytical ability and knowledge capture and use, without increasing users' cognitive load or disrupting their normal way of working. 5. MARLIES - WebIE Supporting the Supply of Manufacturing | A.Univ.-Prof. Dr. Birgit Proell, Johannes Kepler University Linz Sourcing and procurement are mayor tasks of supply chain management, which increasingly benefit from Web technologies in that companies make information on their services retrievable online for the prospective consumers. Web Information extraction technologies constitute the prerequisite to enable the semi-automatic support for finding the best fitting service supplier on the Web. In our presentation we introduce MARLIES, which is targeted on extracting contact data and data on services, e.g., machines and manufacturing processes offered by companies via their heterogeneously designed Web sites. MARLIES implements an ontology-aware knowledge engineering approach based on the text engineering framework GATE. Major challenges are (1) the ontological modelling of the complexly related technical data, (2) the extraction of relations between entities addressing structural aspects, e.g., related entities concealed in nested tables, and (3) ranking of results using various relevance criteria.
1. KnowleSuite: Bringing deep search to end-users with Semantic Technologies | K. Rajaraman and Ang Wee Tiong, Institute for Infocomm Research Digital information available as un/semi-structured data is estimated to comprise of 53% of all available information in enterprises, and little of it makes way into data warehouses. In addition, state-of-the-art search technologies for accessing information from online sources produce highly efficient query issue but are in-effective in delivering targeted results. They provide little support to articulate complex queries which results in poor delivery and indexing of harvested information. We present a generic service-oriented semantic technology infrastructure, called Knowle platform, that uses a combination of text mining, ontologies, and reasoning techniques, and enables visual query navigation and data mining of un/semi-structured, and structured data indexed with OWL-DL ontologies. Knowle has been applied to a variety of domains including Healthcae & Life sciences, Business Intelligence, Patent Landscaping and Campaign management. 2. Demo: Extending Traditional Enterprise Search using Semantic Techniques | Steve Fullerton and Ian Pallen, Solcara Limited Building on the power and flexibility of a federated search tool that enables an asynchronous search approach to internal and external information sources, this demonstration shows how a semantic federating layer, using the OpenRDF Sesame SAIL API, integrates traditional and semantic data sources. 3. The IRF and making search research more relevant to practice | Sabine Ringhofer, IRF The Information Retrieval Facility (IRF) is an international not-for-profit foundation based in Vienna. The IRF is dedicated to the promotion and facilitation of research on large-scale information retrieval and is controlled by a Scientific Board consisting of 13 distinguished scientists in information retrieval and related disciplines, chaired by K. van Rijsbergen. The IRF is a member of the LarKC project consortium, which is building an infrastructure for very large scale inference with a semantic web environment. Another major focus for the IRF is the evaluation of the effectiveness of information access and search systems especially for patents. The IRF initiated and took a leading role in the NIST TREC CHEM 2009 and the CLEF IP 2009 evaluation.The IRF is a membership organization open to both academic and industrial members. A high performance computing facility is operated by the IRF in Vienna, and made available to its members for research purposes. Standard IR packages like Lemur/Indri and Terrier are available. Large datasets are also made available, including MAREC, a corpus of 19 million patent documents and the ClueWeb09 collection of a thousand million web pages. BUSINESS TOOLS AND APPLICATIONS I (Business Intelligence) 1. Sig.ma: to see the Web of data, to foresee its future | Giovanni Tummarello, DERI We present Sig.ma, both a service and an end user application to access the Web of Data as an integrated information space. 2. WEB 2.0 and Semantic Technologies for Open Innovation | Guillermo Alvaro, iSOCO In recent years, Web 2.0 has achieved a key role in the Internet community and applications based on collective intelligence have grown in importance in the enterprise context. Companies are adopting this paradigm for their internal processes in the so-called Enterprise 2.0. The integration of Enterprise 2.0 with semantic technology seems to be essential for its successful adoption with a positive return of investment. On the other hand, the introduction of the Open Innovation model, for which the innovation process should be opened out of the R&D department to all the employees and external actors, requires a technological infrastructure to be supported. In this paper we discuss how the Web 2.0 philosophy and Semantic Technology support Open Innovation and how three big European companies have profited from this new paradigm.
Semantic MediaWiki, the semantic extension of MediaWiki, promises a big potential for a particular class of commercial applications which share common characteristics. Project management is an example of this class of applications where Semantic MediaWiki appears to be well suited to fuse text and data and – as a side effect – to foster collaborative work in a project team.
How can knowledge workers benefit from the semantic web?
The process of building and maintaining thesauri, especially in corporate settings, quite often cannot be justified due to its unfavourable cost-benefit ratio. Combining different approaches to build thesauri like text-mining, machine learning, expert driven thesaurus modelling and linking open data seems to be a promising methodology to overcome this obstacle. This paper gives an overview over a thesaurus building methodology based on Linked
We see an increased interest from telecom providers, the transportation industry and defense integrators in tracking hundreds of thousands to millions of moving objects in real time. Consider fleets of trucks, swarms of airplanes, track data for soldiers on the battlefield, or track data for endangered animals.
Future semantically-enhanced tourist information systems will be very dynamic. They will focus more on individual tourists, their preferences and their current locations. To interpret this data properly these holistic, situation- aware systems will need to respect more external, such as weather or traffic conditions.
Searching for Web Services on the Web today is a hard task. General-purpose search engines are not precise enough and dedicated service portals are too limited in scale. Service-Finder helps in solving this problem. It employs Information Retrieval and Extraction techniques (like crawling, automatic annotation and indexing) as well as Semantic Technologies, to turn the unstructured and distributed information about Web Services into structured data that expresses service semantics with respect to service-domain ontologies. The result is a Web portal (http://demo.service-finder.eu) offering search functionalities to discover services. In comparison with existing solutions, it achieves both a higher recall, because it indexes more than 25,000 services, and a higher precision, thanks to the semantic annotations that enrich the service descriptions. This represents a big advantage for end users, who can easily discover unknown services that fit their need. 4. Practical Applications of Semantic Technologies in the Enterprise | Mike Cataldo, CEO Cambridge Semantics 1. Bottom-up Ontology Design with Semantic Rules | Chris Moran, Tellilink Corporation Designing a system that uses semantic rules requires not only a new way of thinking about the business logic of the system but also the semantics of the system. Semantic rules are usually thought of as an extension of the model—as a way to enable rules for classification to be expressed in an ontology. In practice, this approach creates overly complex and fragile models, and overly simplistic logic.
This paper introduces an approach to building systems with semantic rules that separates system logic from the system model and system data. It enables models to be simple and general; and it enables logic to be applied to a concept in the context of a system or business process. This separation of the model and the logic means that very general models can be tailored for very specific use. To think of it in human terms: a person’s response to “danger” might always be the same—fight or flee—we are very general systems in a sense. But the concept of Danger, and the logic of classifying something as Dangerous, changes constantly depending on the situation. What distinguishes us as a species is that concepts like Danger are not hard-wired into us. The logic that lets us classify a situation as Dangerous is separate from the concept itself and how we respond to it. 2. Semantic Approach to Privacy Protection | Miroslav Blako, Czech Technical University in Prague P3P is a W3C recommendation for expressing service privacy practices in a structured computerreadable form. However, the lack of formal semantics makes it inconvenient for fully computerized matching against privacy preferences (e.g. the APPEL language). SemPref is an interesting approach that defines several possible semantics for P3P using relational database technology. However, it still doesn't have capabilities to detect errors that frequently occur when designing both privacy preferences and service privacy practices. This paper analyses the problem of detecting semantic errors in P3P policies in depth and proposes a solution based on semantic technologies, namely the OWL Web Ontology Language and the SPARQL-DL query language.
3. Design of a Cost Efficient Methodology to Compare Search Methods | Robert Tolksdorf, Freie Universität Berlin Since the advent of the Internet, search engines play a prominent role for the search and discovery of resources. They developed from pure full text indexing mechanisms over indexing systems accounting for the structure of the web, to systems, which produce answers instead of plain resource lists. However, the methods for comparing their result quality remained nearly unchanged over the years and are still based on the information retrieval measures: precision and recall. Even though these measures are well understood, they lack certain important properties when applied to internet search engines. In order to overcome their weaknesses, we present a new, cost efficient approach for comparing search methods based on their results. We used this new method to compare quantitatively the semantic search technology developed by our company with the full text search of a major topic-specific internet portal.
4. Semantic Mashups for the Masses | Thorsten Liebig, Derivo GmbH This talk introduces Jigs4OWL, a semantic mashup framework which offers a web-enabled semantic infrastructure while hiding all non-relevant technical details. It allows to easily build web browser applications which aggregate, visualize, and mashup semantic content.
The framework consists of a server which hosts semantic data uploaded by end users. The framework also delivers a collection of graphical widget templates for dealing with this content. A semantic application is build by customizing the templates and combining it with further widgets (such as Google Maps) or other data sources. This allows to develop consumer and business mashups with a few lines of code and without the burden of operating a complex application back-end. The framework can be either driven by a company as an internal or external service or employed on-demand from a framework service provider. A public accessible instance of the latter is online available for evaluation at www.jigs4owl.com. In this talk we introduce the mashup framework and discuss the wide range of application scenarios. We will also exemplarily show how to adapt and configure it at one‘s own needs. USING TECHNOLOGIES II (Linked Open Data) 1. Reasoning with a Billion of Linked Data Facts | Atanas Kiryakov, Ontotext AD LDSR is a collection of linked data datasets, selected and refined to present a reason-able view to the center of the LOD cloud. LDSR consists of 915 million explicit statements and includes DBPedia, Freebase, Geonames, Wordnet, CIA Factbook, lingvoj, UMBEL, and the ontologies, used by them (e.g. SKOS, FOAF, and DC).
We loaded LDSR in the OWLIM semantic repository, inferring 655 million statements that have been indexed along the explicit ones (OWL dialect similar to OWL 2 RL was used). Although OWLIM performs complete forward-chaining, it does not materialize all the results – while the total number of all indexed statements is 1.57 billion, the number of retrievable statements is 3.02 billion. LDSR is available for search, exploration, and query evaluation at http://www.ontotext.com/ldsr/. LDSR and its front-ends develop constantly in order to present the best possible way for someone to comprehend and start using the web of linked data. 2. Linked data for Personal Information Management | Girts Niedra, TenForce It proves to be very difficult to manage the personal and public information on the web, resulting from the magnitude of portals and users requesting the same up-to-date information. In this paper the author is proposing a community - driven project, called LinkedPIM, which uses semantic technologies and open subscription protocols to make delivery of up-to-date Personal Information Management artefacts easy and reliable.
3. SKOS Extensions for the EUROVOC Thesaurus | Johan De Smedt, TenForce The EUROVOC Thesaurus is now available in RDF. For this purpose, EUROVOC has been modeled using OWL. The resulting EUROVOC ontology is a SKOS and SKOS-XL extension. The paper and the presentation details how thesaurus specific artifacts of EUROVOC are represented complying to SKOS, to SKOS-XL and to the EUROVOC specific ontology.
Attention is given to the following topics: (i) thesaurus organization in domain and micro-thesaurus; (ii) thesaurus release identification; (iii) management of the vocabulary, different equivalence relationships and permuted labels; (iv) sub-types of concepts; (v) usage of 'notes' and notes formatting (using XHTML + RDFa); (vi) documenting rules applied for generating SKOS and SKOS-XL compliant representations. |