A Knowledge Based Approach to Identification of Serial Verb Construction in Chinese-to-Korean Machine Translation System

In Chinese language processing, the recognition and analysis for serial verb constructions (SVCs) is a fascinating research topic. Chinese language researchers each may have a different definition and interpretation of SVC since the structure of SVC makes Chinese unique to other languages and contains complex semantic and pragmatic information. This paper proposes a formal definition of SVC and a knowledge based approach for the recognition of SVCs, which is adopted in TOTAL-CK, a transfer-based MT system from Chinese to Korean. The recognition process is carried out in two stages: the analysis stage classifies SVCs into general categories, and the transfer stage performs further classification for Korean transfer. Some evaluation result for each stage was also given with statistics of each category of SVCs