Learning to Parse Spontaneous Speech

Abstract

We describe and experimentally evaluate a system, FeasPar, that learns parsing spontaneous speech. To train and run FeasPar (Feature Structure Parser), only limited handmodeled knowledge is required. The mbox{FeasPar} architecture consists of neural networks and a search. The networks spilt the incoming sentence into chunks, which are labeled with feature values and chunk relations. Then, the search finds the most probable and consistent feature structure. FeasPar is trained, tested and evaluated with the Spontaneous Scheduling Task, and compared with two samples of a handmodeled GLR* parser, developed developed for 4 months and 2 years, respectively. The handmodeling effort for FeasPar is 2 weeks. FeasPar performes better than the GLR* parser developed 4 months in all six comparisons that are made and has a similar performance as the GLR* parser developed for 2 years.