Title :
SEARCHY: An Agent to Personalize Search Results
Author :
Marcialis, Ivan ; De Vita, E.
Author_Institution :
CRS4, Pulaca
Abstract :
User´s behaviour in browsing sessions is a valuable source of information useful to analyze user interests and personalize the human-computer interaction during information seeking tasks. In this paper we present a simple way to machine-learn user profile and employ it to improve Web searches. Searchy is a personal agent, embedded in the Web browser, able to sense user´s tastes, to assist her/him during query formulation phase and, finally, to filter and sort the results according to the user´s needs. Searchy uses the DART P2P network to store user´s profiles in order to enable collaborative recommendations.
Keywords :
information filtering; online front-ends; peer-to-peer computing; query formulation; software agents; P2P network; Searchy; Web browser; Web searches; browsing sessions; collaborative recommendations; human-computer interaction; information seeking tasks; machine-learn user profile; personal agent; query formulation phase; search results personalization; user behaviour; user profiles; Algorithm design and analysis; Information filtering; Information filters; Information resources; Information retrieval; Search engines; Software agents; Web and internet services; Web pages; Web search; Cooperative systems; feedback; information retrieval; user modeling;
Conference_Titel :
Internet and Web Applications and Services, 2008. ICIW '08. Third International Conference on
Conference_Location :
Athens
Print_ISBN :
978-0-7695-3163-2
Electronic_ISBN :
978-0-7695-3163-2
DOI :
10.1109/ICIW.2008.86
