Semi-automatic identification of danish discourse deictics
Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › fagfællebedømt
In this paper we present an algorithm for the (semi-)automatic identification of anaphors whose antecedents are verbal phrases, clauses or discourse segments in Danish Dialogues. Although these anaphors are quite frequent, especially in conversations, they are usually been neglected in computational linguistics. The algorithm we propose contains defeasible rules for distinguishing these anaphors from those who have individual nominals as antecedents. The rules have been identified by looking at the occurrences of these types of anaphor in the transcriptions of two dialogue collections. The algorithm has been manually tested on four Danish dialogues and the obtained results have been evaluated.
Originalsprog | Engelsk |
---|---|
Titel | Text, Speech and Dialogue - 4th International Conference, TSD 2001, Proceedings |
Redaktører | Vaclav Matousek, Pavel Mautner, Roman Moucek, Karel Tauser |
Antal sider | 7 |
Forlag | Springer Verlag |
Publikationsdato | 2001 |
Sider | 396-402 |
ISBN (Trykt) | 9783540425571 |
DOI | |
Status | Udgivet - 2001 |
Begivenhed | 4th International Conference on Text, Speech and Dialogue, TSD 2001 - Zelezna Ruda, Tjekkiet Varighed: 11 sep. 2001 → 13 sep. 2001 |
Konference
Konference | 4th International Conference on Text, Speech and Dialogue, TSD 2001 |
---|---|
Land | Tjekkiet |
By | Zelezna Ruda |
Periode | 11/09/2001 → 13/09/2001 |
Navn | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Vol/bind | 2166 |
ISSN | 0302-9743 |
Bibliografisk note
Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 2001.
ID: 301814657