Abstract | ||
---|---|---|
This paper presents a shallow dialogue analysis model, aimed at human-human dialogues in the context of staff or business meetings. Four components of the model are defined, and several machine learning techniques are used to extract features from dialogue transcripts: maximum entropy classifiers for dialogue acts, latent semantic analysis for topic segmentation, or decision tree classifiers for discourse markers. A rule-based approach is proposed for solving cross-modal references to meeting documents. The methods are trained and evaluated thanks to a common data set and annotation format. The integration of the components into an automated shallow dialogue parser opens the way to multimodal meeting processing and retrieval applications. |
Year | DOI | Venue |
---|---|---|
2004 | 10.1007/978-3-540-30568-2_24 | MLMI |
Keywords | Field | DocType |
annotation format,meeting document,meeting processing,business meeting,latent semantic analysis,dialogue act,shallow dialogue analysis model,human-human dialogue,dialogue transcript,shallow dialogue processing,automated shallow dialogue parser,maximum entropy,machine learning,rule based,discourse marker,decision tree classifier | Decision tree,Annotation,Segmentation,Computer science,Natural language processing,Artificial intelligence,Parsing,Principle of maximum entropy,Latent semantic analysis,Brown Corpus,Machine learning,Discourse marker | Conference |
Volume | ISSN | ISBN |
3361 | 0302-9743 | 3-540-24509-X |
Citations | PageRank | References |
4 | 0.48 | 15 |
Authors | ||
5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Andrei Popescu-Belis | 1 | 573 | 64.13 |
Alexander Clark | 2 | 43 | 4.81 |
Maria Georgescul | 3 | 66 | 7.23 |
Denis Lalanne | 4 | 836 | 61.09 |
Sandrine Zufferey | 5 | 49 | 4.98 |