An Event Model Adapted for Analysis of Verb Sense in Natural Language

摘要

The verb and its dominated semantic elements constitute an event. The event is one of the basic elements constituting natural language. This paper tries to analyze the verb sense under the level of event. First, the relation of verb and event is analyzed, and an event model adapted for representing the event connotation, the extension and the relation between them is presented. Second, the knowledge of event in natural language is abstracted and classified based on the event model, the static event which described the abstract knowledge of event is given, and the dynamic event which described the verb-noun chunk in the environment of actual language is given too. Third, this knowledge is adapted for describing the verb-noun chunks, and the algorithm adapted for analysis of verb sense is given. Finally, the result of experiment shows that this proposed algorithm is feasible and effective.