The Construction of the Semantic Collocation Database of Verb-Complement Structure in Modern Chinese based on a Large-scale Chinese Chunkbank

In natural language understanding, the marked verb-complement structure can be identified by the complement markers such as $\unicode{x5F97}$(an auxiliary word). However, it is difficult to correctly identify the unmarked verb-complement structure and label the semantic categories expressed by it. The current researches lack a semantic collocation database for the full picture of the verb-complement structure. This paper builds a semantic collocation database based on a large-scale Chinese chunkbank for the verb-complement structure. Firstly, we summarized a comprehensive semantic classification system of the verb-complement structure; moreover, we subclassified the major categories and added the extensional meaning of complement, including seven major categories and twenty-five subcategories. Secondly, we formalized the verb-complement structure and wrote twenty-seven types of formal expressions. Then, we extracted the verb-complement collocation from the large-scale Chinese chunkbank according to the formal expressions, disambiguated, and semantically subclassified the extraction results. Finally, a semantic collocation database containing about 180,000 verb-complement collocations is constructed and statistically analyzed. This paper provides a relevant research paradigm for constructing the semantic collocation database and the proper structure and rich semantic knowledge for natural language understanding.