Using integer linear programming for detecting speech disfluencies

Authors:
Kallirroi Georgila
Affiliations:
University of Southern California, Marina del Rey, CA
Venue:
NAACL-Short '09 Proceedings of Human Language Technologies: The 2009 Annual Conference of the North American Chapter of the Association for Computational Linguistics, Companion Volume: Short Papers
Year:
2009

Citing 6
Cited 1

Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data

ICML '01 Proceedings of the Eighteenth International Conference on Machine Learning
Maximum entropy models for natural language ambiguity resolution

Maximum entropy models for natural language ambiguity resolution
Speech repairs, intonational phrases, and discourse markers: modeling speakers' utterances in spoken dialogue

Computational Linguistics
Statistical language modeling for speech disfluencies

ICASSP '96 Proceedings of the Acoustics, Speech, and Signal Processing, 1996. on Conference Proceedings., 1996 IEEE International Conference - Volume 01
Global inference for sentence compression an integer linear programming approach

Journal of Artificial Intelligence Research
Enriching speech recognition with automatic detection of sentence boundaries and disfluencies

IEEE Transactions on Audio, Speech, and Language Processing

Cross-domain speech disfluency detection

SIGDIAL '10 Proceedings of the 11th Annual Meeting of the Special Interest Group on Discourse and Dialogue

Quantified Score

Hi-index	0.00

Visualization

Abstract

We present a novel two-stage technique for detecting speech disfluencies based on Integer Linear Programming (ILP). In the first stage we use state-of-the-art models for speech disfluency detection, in particular, hidden-event language models, maximum entropy models and conditional random fields. During testing each model proposes possible disfluency labels which are then assessed in the presence of local and global constraints using ILP. Our experimental results show that by using ILP we can improve the performance of our models with negligible cost in processing time. The less training data is available the larger the improvement due to ILP.