Skip to Main content Skip to Navigation

Recherche d'information sur Internet par algorithmes évolutionnaires

Abstract : In this thesis, we present the problem of information search on Internet and more generally of strategic watch. We generally notice that it is necessary to spend a long time to analyze the results provided by traditional search engines in order to obtain good answers. In this thesis, we have developed an automatic search tool based on evolutionary search strategies. This tool explores Web pages starting from the results provided by traditional search engines (like Google, Altavista ...). Several methods of optimization have been compared: a genetic approach, a population of ants and a tabu search. The search effort is also parallelized and can be distributed on several distant computers in order to maximize the resources available for the execution of this task and in order to user a large scale parallel architecture. Finally, we propose a display system of search engine results based on the property of flock of agents in order to help the users to better understanding the results returned by the engine and to decrease the time necessary to analyse these results.
Document type :
Complete list of metadata
Contributor : Fabien Picarougne Connect in order to contact the contributor
Submitted on : Sunday, March 20, 2005 - 11:20:58 PM
Last modification on : Monday, October 19, 2020 - 10:54:50 AM
Long-term archiving on: : Friday, April 2, 2010 - 8:52:44 PM


  • HAL Id : tel-00008013, version 1


Fabien Picarougne. Recherche d'information sur Internet par algorithmes évolutionnaires. Autre [cs.OH]. Université François Rabelais, 2004. Français. ⟨tel-00008013⟩



Record views


Files downloads