Research on Methods of Chinese Sentence Semantic Matching Based on Conceptual Graph with Weight
-
Abstract
A semantic matching method is presented, which is based on conceptual graph with weight. Chinese statement information can transform conceptual graph according to some rules, and each concept nodes are assigned a weight.The word similarity is calaclated through semontatic library, then the similarity between the nodes of each concept is calculated with the node weights, calculated as the similarity of the overall conceptual graph, to achieve the purpose of Chinese Information semantic matching. Compared with traditional keyword semantic matching method, the proposed method is better for the match effect of Synonyms Word and Homographs Word. Experimental results show that the method can effectively achieve semantic matching of Chinese statement, and semantic matching accuracy reached 60%.
-
-