Skip to Main content Skip to Navigation
Journal articles

Automatically Characterizing Linguistic' Salience Using Readers Feedback

Jean-Yves Delort 1, *
* Corresponding author
1 LIRMM/HE - Hors Équipe
LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier
Abstract : Salience is an important characteristic of information influencing users' cognitive and emotional states. For example, salient parts of a document are those that will be considered moving or provoking by readers. This article studies the concept of salience and its specific meanings in linguistics. Then it analyses the main difficulties of content-based techniques for automatic identification of salient passages in a document. A new, context-based method for overcoming these difficulties is subsequently presented. Our method identifies passages that readers have reacted to by analyzing their textual feedback. Our experimentation revealed that it is effective and can be broadly used.
Document type :
Journal articles
Complete list of metadata
Contributor : Jocelyne Nanard <>
Submitted on : Friday, February 27, 2009 - 10:30:09 AM
Last modification on : Thursday, May 24, 2018 - 3:59:21 PM


  • HAL Id : lirmm-00364752, version 1



Jean-Yves Delort. Automatically Characterizing Linguistic' Salience Using Readers Feedback. Journal of Digital Information, Texas Digital Library, 2009, 10 (1), pp.22. ⟨lirmm-00364752⟩



Record views