Proceedings chapter
Open access

Exploiting Structural Meeting-Specific Features for Topic Segmentation

Published inActes de TALN/RECITAL, p. 15-24
Presented at Toulouse (France)
Publication date2007

In this article we address the task of automatic text structuring into linear and non-overlapping thematic episodes. Our investigation reports on the use of various lexical, acoustic and syntactic features, and makes a comparison of how these features influence performance of automatic topic segmentation. Using datasets containing multi-party meeting transcriptions, we base our experiments on a proven state-of-the-art approach using support vector classification.

Citation (ISO format)
GEORGESCUL, Maria, CLARK, Alexander, ARMSTRONG, Susan. Exploiting Structural Meeting-Specific Features for Topic Segmentation. In: Actes de TALN/RECITAL. Toulouse (France). [s.l.] : [s.n.], 2007. p. 15–24.
Main files (1)
Proceedings chapter
  • PID : unige:3463

Technical informations

Creation10/02/2009 9:28:58 AM
First validation10/02/2009 9:28:58 AM
Update time03/14/2023 3:14:53 PM
Status update03/14/2023 3:14:53 PM
Last indexation02/12/2024 6:13:21 PM
All rights reserved by Archive ouverte UNIGE and the University of GenevaunigeBlack