Research On Ontology For Supporting Chinese Sentence Deep Analysis

Natural language understanding is a challenging problem in the field of artificial intelligence, it aims to make the machine really understand human language and communcate with human without difficult. While the natural language has conventional, multi-valued, thumbnail, implied and metaphorical features, which make it difficult for computer to understanding natural language. The reason that natural language understanding problem has not been properly resolved is the lack of knowledge in the process of computers to understand natural language. As a knowledge representation founction, ontology give a formal definition of concepts and the relationship between concepts in a certain field and provide knowledge sharing and reuse. Based on ontology and from the common sense perspective, this paper reaserch the knowledge needed in the process of natural language understanding and how to acquire the knowledge.The contributions of this dissertation are as follows:First of all, this paper proposed a knowledge model of event. In this paper, we take a sentence framework in modern Chinese as an event. Events were expressed as a7-tuple in knowledge model. That is E=(CW, L, P, LCS, PK, EK, R). In this case, the event model contains not only the language contents but also the necessary knowledge to understand the event, so that we can use this model for event reasoning and understanding.Secondly, this paper proposed a method of event knowledge analysis. In this work, we proposed a method of event knowledge analysis in order to analysize preconditions and effects of events from multiple perspectives. In the process of analysis we follow three phases:fundamental analysis, knowledge supplement, and knowledge optimization. We summarized four common sense knowledge dimensions to solve the problem of comprehensiveness and accuracy in analysis, which are physiological, psychological, social and physical. The method of role specification is proposed in the knowledge supplement phase. Experiments show that this method is promising for acquiring comprehensive and accurate commonsense event knowledge.Finally, we developed an ontology management tool of events. This tool not only allows the knowledge engineer to add, modify and query events knowledge, but also modifiy event structure convenient. In Order to guarantee the security of the knowledge base, different type of user rights were set, thus users can only perform operations within his/her authority.