Combining discriminative re-ranking and co-training for parsing Mandarin speech transcripts

Discriminative reranking has been able to significantly improve parsing performance, and co-training has proven to be an effective weakly supervised learning algorithm to bootstrap parsers from a small in-domain seed labeled corpus using a large amount of unlabeled in-domain data. In this paper, we present systematic investigations on combining discriminative reranking and co-training, including co-training reranked parsers and co-training rerankers. We show that combining discriminative reranking and co-training could im

• Type: Conference
• Year: 2009
• Where: ICASSP
• Authors: Wen Wang
• Keywords: Co-training, Co-training Reranked Parsers, Discriminative Reranking, ICASSP 2009, Signal Processing