Cross-lingual query post-translation extension method based on fully weighted positive and negative patterns

Abstract

The present invention provides a cross-language query post-translation consequent extension method based on complete weighted positive and negative mode. The method comprises: firstly, translating a source language query into a target language query and searching a target language document for the target language query, and extracting the preliminarily searched document and constructing a preliminarily searched related document set after related determination of a user; then, by adopting a positive and negative association mode mining technology based on a support-association-enhancement-confidence evaluation framework and oriented to cross-language query extension, mining the preliminarily searched related document set for a feature word positive and negative association rule mode containing a query word item, and constructing a feature word positive and negative association rule library; and extracting a complete weighted positive and negative association rule mode whose rule antecedent is the query word item from the library, and using a positive association rule consequent feature word as a candidate extension word and a negative association rule consequent feature word as a negative extension word, and removing the negative extension word from the candidate extension word to obtain a final consequent extension word for realizing cross-language query post-translation consequent extension. The method provided by the present invention can improve cross-language information retrieval performance and has higher application value and better market prospect.

Description

technical field [0001] The invention belongs to the field of Internet information retrieval, in particular to a cross-language query post-translation extension method based on a fully weighted positive and negative mode, which is suitable for fields such as cross-language information retrieval query expansion. Background technique [0002] Cross-Language Information Retrieval (CLIR) began to receive attention and attention in the late 1990s. Now we are in an era of multilingual networks. Language-diversified Internet resources have become big data information resources, and cross-language retrieval tools with good performance are urgently needed. Therefore, cross-language information retrieval has become an urgent research technology in the field of information retrieval. [0003] Cross-language information retrieval refers to the technology of retrieving information resources in other languages ​​in the form of a query in one language. The language used to express user quer...

AI Technical Summary

Problems solved by technology

[0005] The present invention applies fully weighted positive and negative association pattern mining to post-translation extension of cross-language query, and proposes a post-translation extension method of cross-language query based on fully weighted positive and negative patterns, which is applied to the field of cross-language information retrieval and can solve the problem of The long-standing problems of query subject drift and word mismatch in cross-language information retrieval can improve the performance of cross-language information retrieval, and can also be applied to cross-language search engines to improve the retrieval performance of search engines such as recall rate and precision rate

Images