Title :
A rule-based method for Chinese punctuations processing in sentences segmentation
Author :
Jing Wang ; Yun Zhu ; Yaohong Jin
Author_Institution :
Inst. of Chinese Inf. Process., Beijing Normal Univ., Beijing, China
Abstract :
In this paper, a rule-based sentence segmentation system is proposed. We studied the usage and function of Chinese punctuation marks, and classified them into 4 categories. According to whether punctuation can split a sentence, we tagged it with a label SST or un-SST. Experiments were conducted on 4 different kinds of corpus containing 12 kinds of Chinese punctuation marks, and our model achieves a high F-measure over 90% overall. Experiment results show that our approach is effectively for sentence segmentation.
Keywords :
knowledge based systems; natural language processing; Chinese punctuation marks; Chinese punctuations processing; F-measure; label SST; rule-based sentence segmentation system; sentences segmentation; unSST; Educational institutions; Information processing; Natural language processing; Patents; Presses; Semantics; Syntactics; Chinese Punctuation; Rule-Based Method; Sentence Segmentation;
Conference_Titel :
Asian Language Processing (IALP), 2014 International Conference on
Conference_Location :
Kuching
DOI :
10.1109/IALP.2014.6973504
