Method and system for scientific and technical literature retrieval

Abstract

The invention provides a determiner-based method for Chinese scientific and technical literature retrieval. The method comprises the following steps that firstly, the semantic relevancy between feature vectors of retrieved entries and feature vectors of all scientific and technical literature names in a data set is calculated; then based on the Chinese character 'de' namely 'of', the retrieved entries are subjected to description, so that the scientific and technical literature names with prefixes being the same as those of the retrieved entries are found out, the semantic relevancy corresponding to the scientific and technical literature names is corrected, and finally a plurality of scientific and technical literatures with high semantic relevancy corresponding to the names of the scientific and technical literatures are used as retrieval results. According to the method, the semantic relevancy between the retrieved entries and the names of the scientific and technical literatures is considered and combined with relevance between words in the Chinese grammar, and therefore the retrieval efficiency of Chinese scientific and technical literatures is improved.

Description

Technical field [0001] The present invention relates to the fields of information retrieval and data mining, in particular to a retrieval method for scientific and technological documents. Background technique [0002] With the development of information technology and computers, the number of various electronic documents has increased at an unprecedented rate, and electronic documents are gradually replacing traditional paper publications. Electronic document retrieval has become an effective way to obtain information. [0003] Existing electronic document retrieval methods are usually implemented based on statistical word frequency. When entering search keywords, the search results are sorted according to the frequency of the counted keywords appearing in the electronic file. These methods are all mechanically performed statistics and matching, and ignore the actual semantic meaning of the word itself and the semantic relevance between words, and they do not combine the languag...

AI Technical Summary

Problems solved by technology

This kind of method is statistical and matching mechanically, but ignores the actual semantic meaning of the word itself and the semantic correlation between words, and does not combine the language habits of various languages, the retrieval effect not ideal

Especially for scientific and technological literature, some common theories and method terms appear frequently, but these terms with high word frequency cannot well characterize the characteristics of the literature, so the retrieval efficiency of scientific and technological literature is not as good as that of statistical word frequency. not tall

Images