en
Proceedings chapter
Open access
English

Contrasting the Automatic Identification of Two Discourse Markers in Multiparty Dialogues

Published inProceedings of SIGDIAL 2007 (8th SIGdial Workshop on Discourse and Dialogue), Editors ACL - Association for Computational Linguistics, p. 10-17
Presented at Antwerp (Belgium), September
Publication date2007
Abstract

The identification of occurrences of like and well that serve as discourse markers (DMs) is a classification problem which is studied here on a corpus of dialogue transcripts with more than 4,000 occurrences of each item. Decision trees using item-specific lexical, prosodic, positional and sociolinguistic features are trained using the C4.5 method. The results demonstrate improvement over past experiments, reaching the same range as inter-annotator agreement scores. DM identification appears to benefit from itemspecific classifiers, which perform better than general purpose ones, thanks to the differentiated use of lexical features.

Citation (ISO format)
POPESCU-BELIS, Andréi, ZUFFEREY, Sandrine. Contrasting the Automatic Identification of Two Discourse Markers in Multiparty Dialogues. In: Proceedings of SIGDIAL 2007 (8th SIGdial Workshop on Discourse and Dialogue). Antwerp (Belgium). [s.l.] : [s.n.], 2007. p. 10–17.
Main files (1)
Proceedings chapter
accessLevelPublic
Identifiers
  • PID : unige:3487
497views
243downloads

Technical informations

Creation10/02/2009 9:31:19 AM
First validation10/02/2009 9:31:19 AM
Update time03/14/2023 3:14:59 PM
Status update03/14/2023 3:14:59 PM
Last indexation02/12/2024 6:13:48 PM
All rights reserved by Archive ouverte UNIGE and the University of GenevaunigeBlack