"Adaptive Information Integration"

11/05/04: 10:30 AM
11th Floor Large Conference Room
Host: Patrick Pantel, schedule

Abstract: The availability of structured and semi-structured information sources on the web has recently lead to significant interest in query processing frameworks that can integrate data sources available on the web. An important impediment to scaling up information integration frameworks to large scale applications has been the fact that the autonomous and decentralized nature of the data sources constrains the mediators to operate with very little information about the structure, scope, and contents of the information sources they are trying to integrate. To be effective, the mediator needs to be aware of, and be adaptive to: the capabilities of the sources, the data spread characteristics among sources (e.g. coverage of the individual sources, overlaps between them), as well as the often imprecisely expressed query needs of the user. In our recent work, we have made significant progress on designing and developing such adaptive information integration frameworks--including 1. BibFinder, a bibliography mediator that is sensitive to the coverage and overlap statistics over structured data, and supports multi-objective query processing. 2. COSCO, a collection selection tool which combines relevance and overlap analysis for unstructured (text) data. 3. AIMQ, a novel framework for supporting imprecise queries over structured and semi-structured sources. In this talk, I will describe the challenges of adaptive integration as well as the results of our recent research.

About Subbarao Kambhampati: Subbarao Kambhampati is a professor of Computer Science and Engineering at Arizona State University. His research interests are in AI (Automated planning, scheduling, CSP etc.) and Information Integration. He is a 1994 NSF Young Investigator and a 2004 AAAI Fellow. Kambhampati is an Associate Editor of JAIR, co-chaired the 2000 International Planning Conference, and will be co-chairing 2005 National Conference on AI.

Last updated: Mon Jun 19 17:44:06 2006