Abstract | ||
---|---|---|
This article discusses the detection of dis- course markers (DM) in dialog transcriptions, by human annotators and by automated means. After a theoretical discussion of the definition of DMs and their relevance to natu- ral language processing, we focus on the role of like as a DM. Results from experiments with human annotators show that detection of DMs is a difficult but reliable task, which re- quires prosodic information from soundtracks. Then, several types of features are defined for automatic disambiguation of like: colloca- tions, part-of-speech tags and duration-based features. Decision-tree learning shows that for like, nearly 70% precision can be reached, with near 100% recall, mainly using colloca- tion filters. Similar results hold for well, with about 91% precision at 100% recall. |
Year | Venue | Field |
---|---|---|
2004 | SIGDIAL Workshop | Dialog box,Transcription (linguistics),Computer science,Speech recognition,Natural language processing,Artificial intelligence,Recall,Discourse marker,Collocation |
DocType | Citations | PageRank |
Conference | 7 | 0.68 |
References | Authors | |
2 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Sandrine Zufferey | 1 | 49 | 4.98 |
Andrei Popescu-Belis | 2 | 573 | 64.13 |