Subcategorization Acquisition and Evaluation for Chinese Verbs.pdfVIP

Subcategorization Acquisition and Evaluation for Chinese Verbs Xiwu Han, Tiejun Zhao, Haoliang Qi, Hao Yu Department of Computer Science, Harbin Institute of Technology, 150001 Harbin, China {hxw, tjzhao, qhl, yh}@ and semantic generalizabilities of the consequent Abstract outputs are not in such a harmony that satisfies the description granularity for SCF (Han and This paper describes the technology and an ex- Zhao 2004). The only auto-acquisition work for periment of subcategorization acquisition for Chinese SCF made by (Han and Zhao 2004) de- Chinese verbs. The SCF hypotheses are gener- scribes the predefinition of 152 general frames ated by means of linguistic heuristic information for all verbs in Chinese, but that experiment is and filtered via statistical methods. Evaluation not based on real corpus. After observing and on the acquisition of 20 multi-pattern verbs shows that our experiment achieved the similar analyzing quantity of subcategory phenomena in precision and recall with former researches. Be- real Chinese corpus in the People’s Daily sides, simple application of the acquired lexicon (Jan.~June, 1998), we removed from Han to a PCFG parser indicates great potentialities of Zhao’s predefinition 15 SCFs that are