A Semantic Query Expansion Method Based on Domain Knowledge

Abstract

The invention discloses a semantic query extension method based on domain knowledge, which includes: constructing domain knowledge based on concept expression and knowledge tree system; performing preliminary semantic analysis on user query phrases to form a list of semantic items; utilizing preliminary semantic analysis Based on domain knowledge, construct a semantic graph with extended types and extended weights; calculate the semantic distance between each vertex in the semantic graph and the initial vertex; determine the possible value of each item in the semantic item list according to the semantic distance Extended items; finally relying on the logical relationship of "AND" and "OR" to merge all expandable items to obtain a set of semantic items representing the user's query intention, and submit it to the search system for searching. The invention has short calculation time, fully utilizes domain knowledge, and has a clear semantic relationship between the newly added extended semantic item and the original query phrase, and can effectively improve the recall rate and precision rate of the search system.

Description

technical field [0001] The invention belongs to the field of computers, relates to the application technology of natural language processing, specifically aims at network search engines, and proposes a semantic query expansion method based on domain knowledge. Background technique [0002] With the development of the Internet, the utilization rate of search engines in the United States has reached 90%, and it also reached 72.4% in China in 2008. Internet users are more and more inclined to obtain the resources they need through searching. Internet public information services represented by search engines are developing vigorously. In the case that the general search engine market is monopolized by famous companies such as Google, Microsoft, Yahoo, and Baidu, new search algorithms and technologies are constantly emerging, and personalized, specialized, and in-depth vertical searches are becoming another new step in the development of search engines. The bright spot is the em...

AI Technical Summary

Problems solved by technology

[0008] The purpose of the present invention is to solve the problem that when users use a search engine, they input less query words and cannot fully express the user's search intention, so the recall rate and accuracy rate of search results cannot meet the needs of users. The present invention aims at natural language Semantic correlation characteristics, propose a semantic query extension method based on domain knowledge

Images