Paper: Using Integer Linear Programming for Detecting Speech Disfluencies

ACL ID N09-2028
Title Using Integer Linear Programming for Detecting Speech Disfluencies
Venue Human Language Technologies
Session Short Paper
Year 2009
Authors

We present a novel two-stage technique for de- tecting speech disfluencies based on Integer Linear Programming (ILP). In the first stage we use state-of-the-art models for speech dis- fluency detection, in particular, hidden-event language models, maximum entropy models and conditional random fields. During testing each model proposes possible disfluency la- bels which are then assessed in the presence of local and global constraints using ILP. Our ex- perimental 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 im- provement due to ILP.