Skip to Main content Skip to Navigation

Etude et réalisation d'un système d'extraction de connaissances à partir de textes

Abstract : This PhD dissertation relates to the problems of knowledge extraction from texts, or text mining (TM). It is applied to the text analysis, the datamining process itself, and the interpretation of the elements of knowledge extracted. A system of knowledge extraction necessary to analyse the texts according to their contents is studied. The methods of datamining used are: frequent itemset levelwise search and association rule extraction. The definition of the process of TM and its main characteristics is done. A study of a number of quality measures attached to the rules is carried out. It is shown how far these quality measures can help the interpretation of the extracted rules. The use of a knowledge model comes to support this approach. It is shown, by the definition of a likelihood probability measure, the significance to discover new knowledge by discarding knowledge already described in the domain model. The rules can be used to enrich the knowledge model of the selected domain. This dissertation includes the implementation of the TAMIS system: "Text Analysis by Mining Interesting ruleS" and an experiment on a real-world text corpus holding on molecular biology.
Document type :
File URL :
Complete list of metadatas
Contributor : Thèses Ul <>
Submitted on : Thursday, March 29, 2018 - 10:44:24 AM
Last modification on : Tuesday, August 25, 2020 - 1:59:12 PM


  • HAL Id : tel-01746763, version 1



Hacène Cherfi. Etude et réalisation d'un système d'extraction de connaissances à partir de textes. Autre [cs.OH]. Université Henri Poincaré - Nancy 1, 2004. Français. ⟨NNT : 2004NAN10164⟩. ⟨tel-01746763⟩



Record views