Improving PCFG Chinese Parsing with Context-Dependent Probability Re-estimation
Yu‐Ming Hsieh, Ming-Hong Bai, Jason S. Chang, Keh-Jiann Chen · 2012
Selecting the best structure from several ambiguous structures produced by a syntactic parser is a challenging issue. The quality of the solution depends on the precision of the structure evaluation methods. In this paper, we propose a general model (context-dependent probability re-estimation model, CDM) to enhance the structure probabilities estimation. Compared with using rule probabilities only, the CDM has the advantage of an effective, flexible, and broader range of contexturefeature selection. We conduct experiments on the CDM parsing model by using Sinica Chinese Treebank. The results show that our proposed model significantly outperforms the baseline parser and the open source Berkeley statistical parser. More importantly, we demonstrate that the basic framework of the parsing model does not need to be changed, and the proposed re-estimation functions will adjust the probability estimation for every particular structure, and obtaining the better parsing results. 1