Towards an emergent taxonomy approach for adaptive profiling

  • Authors:

  • Sylvain Videau;Sylvain Lemouzy;Valérie Camps;Pierre Glize

  • Affiliations:

  • IRIT, Paul Sabatier University, Toulouse cedex 09, France;IRIT, Paul Sabatier University, Toulouse cedex 09, France;IRIT, Paul Sabatier University, Toulouse cedex 09, France;IRIT, Paul Sabatier University, Toulouse cedex 09, France

  • Venue:
  • SOCASE'08 Proceedings of the 2008 AAMAS international conference on Service-oriented computing: agents, semantics, and engineering
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

The omnipresence of data processing and mobile telephonyin our life (computers, PDA, GSM, GPS...) along with the evolution ofwireless technologies opens the door towards new habits. To avoid beingsubmerged by too much information it is necessary to equip each electroniccomponent present in user's daily life with capacities to take intoaccount his needs according to his actions, to assist him while learningand anticipating on his behavior in the most autonomous way. Personalizationis clearly situated in this objective; it enables a user profileconstruction which has to dynamically evolve. It also has to take into accountnew preferences, needs and interests of this user and to forget oldones. This paper proposes a local, cooperative and real-time multi-agent approach to build adaptive and incremental profiles. First, documentsare sequentially parsed, which leads to the construction of a TemporaryTerminological Network (TTN). This Network is then merged with otherdocument's extracted networks, in order to create a Permanent TerminologicalNetwork (PTN), relevant to the studied collection and used toindex this collection thanks to a clustering approach. Preliminary resultsof the built system are then presented as well as perspectives.