Semantic web seminar - Jerusalem, October 2002

Citation

Frank, R. (2003), "Semantic web seminar - Jerusalem, October 2002", Library Hi Tech News, Vol. 20 No. 2. https://doi.org/10.1108/lhtn.2003.23920bac.001

Semantic web seminar - Jerusalem, October 2002

Rivkah Frank

The one-day "Semantic Web Seminar" held in Jerusalem, Israel on October 31, 2002, was sponsored by the Israel Directorate for the European Union RTD (Research, Technological Development and Demonstration) Programme (ISERD), the Jewish Agency for Israel and the Israel Internet Association. The purpose was to provide direction to participants to successfully compete for EU grants as well as to showcase some of the technology projects that are now being implemented.

The first session was entitled, "Lessons from Israel Participation in the 5FP EC IST Program" (5th Framework Program of the European Commission's Information Society Technologies). The deadline for the 6th Framework Program (FP6) was December 2002, so that all lessons learned from the previous round could be applied when submitting the new proposals.

The keynote address was given by Roberto Cencioni from the Information Society Directorate General (DG) of the European Commission. He was able to give specific tips and insight as to how the Commission evaluates applications. He explained that the Commission was looking for genuine research into system integration and validation – not simply "engineering" tasks. Also the promotion and dissemination of results of the FP6 instruments should be an integral part of the project.

His topic was "Knowledge Technologies, 2002-2006: Scope and Focus in 2003." The IST aim in FP6 is to apply information technologies to meet major societal and economic challenges. The emphasis in FP6 will be on basic, original research. There are two major research themes:

1. 1.

   Semantic-enabled systems and services – which will include proposals for building semantic Webs within and across organizations and/or communities of interest, and smart Web services featuring products such as networked knowledge discovery, multimedia content mining, content-based retrieval across heterogeneous databases, platforms and networks, and information visualization.

2. 2.

   Knowledge-based adaptive systems – which can act on large volumes of dynamic data and information for automated diagnosis and decision support, highly dynamic and time-critical applications, modeling and optimization, and "anytime-anywhere inferencing."

Cencioni stressed that the EU Commission preference within the 6th Framework Program would be the IP's (Integrated Projects) which would be funded for four years for original basic research. He urged the attendees not to create an "artificial" IP. Rather the IP should be an ambitious and progressive endeavor with clearly defined milestones and checkpoints. Ideally there would be 6-12 multi-country participants. The IP proposal should include genuine research work as well as "engineering" tasks, system integration, and validation, i.e. a total system approach. The proposal should also include promotion and dissemination of results, define cooperation and exchanges with related national and international efforts, and describe socio-economic consequences.

Nahum Korda, the Chief Scientist of VirtualSelf, briefly described the GRACE project (Grid Retrieval and Categorization Engine) that was funded by the EEC (European Economic Commission) under the FP5 and also presented a brief review of the development of the Semantic Web and Semantic Grid technology. As we all know, the technological capabilities of the WWW gave rise to a different definition of a document. No longer do we think of a document as a single file, but rather, as a set of related files whose contents can have different formats – text, multi-media, databases, etc. The files are hyperlinked so that the user can access a composite document. Any document in a Web browser can be linked to other documents, creating additional layers. The aim of the "Semantic Web" is to develop effective methods of creating additional layers automatically, without human intervention. To do this, the computer program has to "understand" the content of the documents. RDF (Resourse Description Framework) is a language that allows efficient storage of metadata describing the content of documents in a "machine-understandable" format.

Grid technology is essentially a "virtual organization." It seeks to develop protocols, services and software that will enable institutions, countries and disciplines to work collectively by accessing software, data and other resources as they engage in collaborative problem-solving initiatives. It will support global sharing of resources.

Korda used the term "service-oriented approach" to describe how the semantic grid will function. He said that the consumer in need of a resource will approach an available service provider in a Grid, and will negotiate the terms under which the required resources will be provided. Because of the underlying complexity of the Grid infrastructure, he said that software agents, not humans, will likely handle these negotiations. Now in order for the "consumer" and "provider" to understand each other the data have to be available in machine-understandable format. The Semantic Grid initiative aims to develop effective methods of making this semantic information available. There are three areas that need to be addressed:

1. 1.

   Data/computational level – which includes such issues as security and authorization, naming services, distributed file systems, etc.

2. 2.

   Information level – creation and management of metadata, workflow management, digital rights, etc.

3. 3.

   Knowledge management – solutions for automatic annotation and summarization, dynamic hyperlinking and meta-tagging, etc.

Both Semantic Web and Semantic Grid utilize ontologies. The creation of ontologies is complex and problematic. And relying on ontologies is fraught with problems: the concepts are often ambiguous, multi-lingual documents create added confusion, and the criteria for defining interrelationships are generally quite inconsistent. To address these problems, Korda described a new technology, Automated Idiomatic Representation (AIR), that is being developed by his company to supplement the ontologies. He said that this technology does not require pre-existing ontologies in order to automatically extract information from the content of a document. And it can be combined with pre-existing ontologies and taxonomies and also can function to automate the further development of these ontologies.

Through the GRACE project, VirtualSelf is currently developing a large-scale Pan-European test bed based on the Grid-enabled version of the AIR technology in collaboration with CERN in Switzerland, Telecom Italia and three other academic institutions.

Another project funded under FP5, The Corporate Ontology Grid Consortium (COG), was presented by Joram Borenstein and Aaron Solomon. This project was funded at 1.8 million Euro for an 18 month period. Its focus is to demonstrate the tangible value of applying Grid technologies to industries, such as manufacturing, engineering, information technology, insurance and finance. It aims to synthesize Grids with ontologies in a fully industrial setting. There are four major partners:

1. 1.

   Unicorn Solutions, based in New York and Jerusalem, which specializes in pioneering semantic information architecture with the Unicorn Coherence product.

2. 2.

   Centro Ricerche Fiat of Italy which develops innovative methodologies to increase competitiveness.

3. 3.

   LogicDIS, which is Greece's leading systems integrator and has many international clients such as Kraft, 3M and Pfizer.

4. 4.

   Universitat Innsbruck which is one of the leading academic centers in ontological development and has experience in such EU projects as OntoWeb and OntoKnowledge.

The expected results are:

• •

  an ontology reflecting industrial processes;

• •

  an information Grid ontological system;

• •

  automatically generated transformation scripts;

• •

  ontology mappings;

• •

  active thesaurus for making data dynamic;

• •

  methodology;

• •

  best practices.

The speakers believe that the project results will ultimately have real value for corporations.

Rivkah Frank (rivkahfrank@hotmail.com) is Director, Medwrite Services, Jerusalem, Israel.