Parsing Speech Repair without Specialized Grammar Symbols

This paper describes a parsing model for speech with repairs that makes a clear separation between linguistically meaningful symbols in the grammar and operations specific to speech repair in the operation of the parser. This system builds a model of how unfinished constituents in speech repairs are likely to finish, and finishes them probabilistically with placeholder structure. These modified repair constituents and the restarted replacement constituent are then recognized together in the same way that two coordinated phrases of the same type are recognized.

• Type: Journal
• Year: 2009
• Where: ACL
• Authors: Tim Miller, Luan Nguyen, William Schuler
• Keywords: ACL 2009, Computational Linguistics, Repair, Speech Repairs, Unfinished Constituents