Curricula Vitae

Dr. rer. nat. Elmar Schalück is a consultant for research funding at the University of Bochum. He works on the distribution of funding information in the Internet since 1993, eg. with the first gopher interface for the CORDIS databases. This has been cancelled meanwhile. He is now working with WWW and Java. Since 1996 he is responsible for ELFI, a pilot project on electronical information gathering and dissemination for research funding.

Dr. rer.nat. Christoph G. Thomas is a computer scientist in the Human Computer Interaction Research Division of GMD FIT. His special research interests include Information Brokering, Software Agents, Adaptive User Interfaces, Learning on Demand, and Distributed Information Spaces. He is chairperson of the special interest group "Adaptivity and User Modeling in Interactive Software systems" of the GI - the German Association for Informatics. He belongs to the editorial board of Knowledge-Based Systems, Elsevier, and to the board of special reviewers of the UMUAI (User Modeling and User-Adapted Interaction, Kluwer) journal. He is co-chair of the IUI99 - ACM International Conference on Intelligent User Interfaces, LA, CA.

ELFI - Electronical Research Funding Information System, Information gathering and personalised dissemination for heterogeneous data.

Parallel Session B.1 and Demonstration
(Parallel Session C)

ELFI is a prototype of an information brokering system suited for the domain of research funding. ELFI itself consists of three parts:

1. A set of agent-based internet search engines searching for funding-relevant information units which are available as
   • WWW based pages (eg. http://www.echo.lu)
   • FTP resources (eg. ftp://ftp.cordis.europa.eu)
   • telnet databases
   • and with different types of data (HTML, WinWord, PDF, ...)
2. A set of categorisation tools for the data documents retrieved by the search engines
   • The documents describe objects in the world of research funding, i.e. funding programmes like the EU programmes in the fourth or fifth framework programme. We extract additional meta-information like deadlines, eligibility criteria (eg. required academic degree for fellowships), the research area of the programme (electrical engineering for the JOULE/Thermie programme of the EU, ...), regional restrictions and many more.
   • The funding organisation, contact partners, and funding oriented events like info days are other objects for which meta-information must be extracted from the documents.
3. A personalised information dissemination implemented as a combination of Push and Pull technologies to provide the user with a personal information subset.
   • Each user defines a personal set of profiles describing the selection criteria for interests in funding programmes specific to the user, eg. for a graduate student in the UK that wants to know about funding possiblities for the preparation of his doctorate thesis. Different profiles define a personal filter on the complete information space, restricting the users view to the most interesting parts.
   • The users profile is used for the Push part of information dissemination. Short notification on new or changed entries in the databases are sent to the user.
   • The personalised WWW page which is adapted to the user´s profile is the last part of the data prepration process. The system will notify the user about new or changed information.
   • In its final version, ELFI will observe the user over time and make suggestions to broaden or to narrow the personal information space.
   • ELFI provides a complete search archive with full text and meta information indexes.
   • ELFI is a information brokerage system, leading the user easily to the original data on the original server.
   • Further developments of ELFI may include collaborative filtering techniques to match different user profiles in order to find possible cooperation partners.

Download their Full paper or their Presentation