Extraction and Application of Verb Event Structure Based on Grammatical Knowledge-Base of Contemporary Chinese(GKB)

Mengxiang Wang, Bingxin Li

Abstract: Although the semantic features of words are hidden deep-seated features, they can generally be shown through the grammatical combination of words. This paper mainly relies on the descriptions of some grammatical features of verbs in GKB, and then extracts the event structure of verbs by features’ stacking and filtering. Finally, the authors use those features to construct semantic knowledge base of verbs. Experiments show that the event structure knowledge is useful for machine translation and semantic role labeling.