Research on Domain Knowledge Based Chinese Short Message Understanding Model

With the rapid increase of the short message service in China, the information query technology based on the Chinese natural language is becoming a research hotspot at present. An algorithm of the Chinese natural language understanding based on certain domain knowledge is proposed, and it is applied to a Chinese short message based information query system. The algorithm is divided into three interrelated parts: word segmentation, syntax analysis and construction of SQL command. The domain knowledge is introduced into the word segmentation part, which simplifies the Chinese semantic understanding. In the syntax analysis part, a syntax analysis technology is integrated with a semantic analysis technology by associating one symbol in grammar rule with an entity and a field of an application database. A syntax tree in the semantic database is constructed as a middle format of SQL transformation and a SQL command is formed by searching the syntax tree in priority of the depth of the syntax tree.