Construction semi-automatique d'index de fin de livre

par Touria Ait El Mekki

Thèse de doctorat en Informatique

Sous la direction de Daniel Kayser.

Soutenue en 2004

à Paris 13 .

  • Titre traduit

    ˜ œBuilding back-of-the book indexes


  • Pas de résumé disponible.


  • Résumé

    [ Résumé en anglais] one of the challenges of text mining is to help readers to exploit large documents relying on their content. The table of contents ofers a simple and direct access to document content but it gives only few, short and often abstract titles as acess points. The search based on an index, which lists the document subjects, is easier and potentially richer. However, the use of indexes for precise information retrieval is brilled by the fact that indexes are often of poor quality. We have defined a whole method for constructing indexes. Our system IndDoc is designed as an assistance tool for the document author or editor. Our approach relies on a clear distinction between on one side the stable knowledge base, the index source, which represents the semantic content of the index and on the other side, the various ways in which the index can be presented to the readers. The overall generation derives one or several indexes from the source acording to editing constraints. The third visualisation step gives its final format to the index.

Consulter en bibliothèque

La version de soutenance existe sous forme papier

Informations

  • Détails : 155 p.
  • Notes : Non reproduite sur microfiches
  • Annexes : Bibliogr. p. 147-155

Où se trouve cette thèse ?

  • Bibliothèque : Université Paris 13 (Villetaneuse, Seine-Saint-Denis). Bibliothèque universitaire. Section Sciences.
  • Non disponible pour le PEB
  • Cote : PARIS13 2004AIT
Voir dans le Sudoc, catalogue collectif des bibliothèques de l'enseignement supérieur et de la recherche.