Distributed Search in the Semantic Web

Goal: Our overall objective is to address the issue of automated distributed search in the context of the Semantic Web, where we assume that an agent may have access to a large number of heterogeneous, distributed and possibly ontology-based information sources.

Topics addressed: automated source selection, automated schema mapping, data fusion, knowledge representation, uncertainty, languages for the Semantic Web

Responsible: Umberto Straccia

Team: Henri Avancini, Elena Renda, Leonardo Candela, Raphael Troncy

Related papers:

1. Nottelmann, Henrik and Straccia, Umberto. A probabilistic approach to schema matching. In Proceedings of the 27th European Conference on Information Retrieval Research (ECIR-05), 2005.
  Abstract . Bibitem . Paper
2. Straccia, Umberto. Uncertainty in Description Logics: a Lattice-based Approach. Proceedings of the 5th International Conference on Information Processing and Managment of Uncertainty in Knowledge-Based Systems, (IPMU-04) , 2004.
  Abstract . Bibitem . Paper
3. Straccia, Umberto. Uncertainty in Description Logic Programs. Technical Report 2004-TR-01, ISTI-CNR, Pisa, ITALY, 2004.
  Abstract . Bibitem . Paper
4. Straccia, Umberto. Uncertainty and Description Logic Programs: A Proposal for Expressing Rules and Uncertainty on Top of Ontologies. Technical Report 2004-TR, ISTI-CNR, Pisa, ITALY, 2004.
  Abstract . Bibitem . Paper

Related seminars:

Henrik Nottelman. Information Systems, Institute of Informatics and Interactive Systems,University of Duisburg, Germany.
- A decision-theoretic framework for cost-optimum resource selection
  - ABSTRACT: Resource selection is the task to decide to which libraries a user query should be routed in a federated system, as it is too expensive to query all libraries. In contrast to standard library ranking approaches like CORI, the decision-theoretic framework follows a different approach on a better theoretic foundation: It computes a selection which minimises the overall costs (e.g. retrieval quality, time, money) of the distributed retrieval. Different methods for the most difficult task, estimating the number of relevant documents in the top-ranked documents, have been developed so far, including an approach which integrated CORI. Experiments showed that our framework is compatible with, and often better than CORI.