Chinese Language Analyzing Using Entity Attribute Semantic Knowledge Base

摘要

A Chinese language analyzing method by using entity attribute semantic knowledge base is proposed in this paper. The knowledge base consists of four parts: the situation verb database, the entity database, the attribute database and the attribute value database. The terms in the above four database are organized into a network according to the relationship between entities and attributes. The analyzing procedure can be generalized into three steps. The first step is to segment the given sentence by matching the terms in the knowledge base; mark the possible semantic categories of every word; specify the possible attribute assignment information for each attribute value word according to the matching relationship between the attribute words and attribute value words. The second step is to identify the entities by the attribute naming rules of instance entities; determine the attribute assignment information of the attribute value words based on the marked attribute information. The third step is to determine the subjects of collocation according to slot information of the situation verb stored in the collocation framework; mark those subjects of collocation. The process of analysis of annotation type is highly consistent with the way of people understanding language. The result of each word of the sentence is consistent with the result of the understanding in the human brain. The proposed method reproduced the complete process of language comprehension, which can be well applied to language analysis.