Verb Pattern: A Probabilistic Semantic Representation on Verbs

• DOI: 10.1609/aaai.v30i1.10334
• Publication Date: 2022-06-23T23:52:13Z
• Authors: Wanyun Cui; Xiyou Zhou; Hangyu Lin; Yanghua Xiao; Haixun Wang; Seung-won Hwang; Wei Wang
• Keywords: FrameNet; Reflexive verb

Verbs are important in semantic understanding of natural language. Traditional verb representations, such as FrameNet, PropBank, VerbNet, focus on verbs' roles. These roles too coarse to represent semantics. In this paper, we introduce patterns semantics, that each pattern corresponds a single the verb. First analyze principles for patterns: generality and specificity. Then propose nonparametric model based description length. Experimental results prove high effectiveness patterns. We further apply context-aware conceptualization, show helpful semantic-related tasks.