Text retrieval method based on word vector learning and pattern mining fusion extension

Abstract

The invention provides a text retrieval method based on word vector learning and pattern mining fusion extension. The method comprises the following steps: querying and retrieving a Chinese document set by a user to obtain an initial retrieval document, and performing word embedding semantic learning training on the initial retrieval document to obtain an initial retrieval document word vector set; constructing a pseudo-correlation feedback document set, mining extension words by adopting a rule posterior extension word mining method based on a Copulas function, establishing a rule consequentextension word set, calculating vector cosine similarity between the rule posterior extension words and an original query, and extracting a word vector rule posterior extension word set; calculating the vector cosine similarity between the non-query lexical items and the original query, and extracting a word vector extension word set; and fusing the word vector extension word set and the word vector rule consequent extension word set union set to obtain a final extension word, and combining the final extension word and the original query into a new query to realize query extension. According to the method, query expansion is realized by adopting a mechanism of two times of retrieval and two times of word vector similarity calculation, and the text information retrieval performance is wellimproved.

Description

technical field [0001] The invention relates to a text retrieval method based on fusion and extension of word vector learning and pattern mining, and belongs to the technical field of information retrieval. Background technique [0002] In the current field of information retrieval, there are still problems of query subject drift and word mismatch, which lead to the degradation of information query performance and affect users' access to required information resources. Using query expansion technology in information retrieval can solve the above problems. Query expansion refers to modifying the weight of the original query, or adding other feature words related to the semantics of the original query, to make up for the lack of semantic information caused by the original query being too simple, and to improve The purpose of information retrieval performance. In the past ten years, scholars have researched information retrieval methods based on query expansion from different ...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to propose a text retrieval method based on the fusion and expansion of word vector learning and pattern mining, and the method is used in the field of information retrieval, such as actual Chinese search engines and web information retrieval systems, which can improve and enhance the performance of information retrieval systems Query performance, reducing query topic drift and word mismatch problems in information retrieval

Images