We propose an approach to user model-based information retrieval which uses an evolutionary algorithm to learn fuzzy models of user interests and to dynamically track their changes as the user interacts with the system. The system is ontology-based, in the sense that it considers concepts behind terms instead of simple terms. The approach has been implemented in a real-world prototype newsfeed aggregator with search facilities called IFeed. Experimental results show that our system learns user models effectively. This is proved by both the convergence of the interest degrees contained in the user models population and the increase of the users' activities on the set of proposed documents.

Learning fuzzy models of user interests in a semantic information retrieval system / M. Dragoni, C. da Costa Pereira, A.G.B. Tettamanzi - In: STAIRS 2010 : proceedings of the fifth starting AI researchers' symposium / [a cura di] T. Agotnes. - Amsterdam : IOS press, 2010. - ISBN 9781607506751. - pp. 76-88 (( Intervento presentato al 5. convegno Starting AI Researchers' Symposium (STAIRS) tenutosi a Lisbon, Portugal nel 2010.

Learning fuzzy models of user interests in a semantic information retrieval system

M. Dragoni
Primo
;
C. da Costa Pereira
Secondo
;
A.G.B. Tettamanzi
Ultimo
2010

Abstract

We propose an approach to user model-based information retrieval which uses an evolutionary algorithm to learn fuzzy models of user interests and to dynamically track their changes as the user interacts with the system. The system is ontology-based, in the sense that it considers concepts behind terms instead of simple terms. The approach has been implemented in a real-world prototype newsfeed aggregator with search facilities called IFeed. Experimental results show that our system learns user models effectively. This is proved by both the convergence of the interest degrees contained in the user models population and the increase of the users' activities on the set of proposed documents.
Settore INF/01 - Informatica
2010
ECCAI
Book Part (author)
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2434/152410
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? 0
social impact