A Uighur language agricultural technology term recognition method

Abstract

The invention discloses a Uyghur character agricultural technology term recognition method, and relates to the technical field of computer application. According to the method, the character string frequency and the C_value value of words of linguistic data are counted in Uyghur character linguistic data; the words corresponding to the C_value value meeting the C_value value threshold are selected and are used as anchor point candidate terms; the statistics features of the anchor point candidate terms are counted; the part-of-speech tagging and the stem and suffix splitting are performed on all words in the linguistic data to obtain language features; a finite state automata is used for integrating the statistics features and the language features to construct a state transfer matrix; the automatic recognition of the agricultural technology term under the control of the finite state automata is realized. The method has the advantages that the accuracy of the Uyghur character agricultural field technology term recognition is improved by 4 percent; the recall rate is improved by about 3 percent; the Uyghur character agricultural field technology term recognition blank is filled.

Description

technical field [0001] The invention relates to the field of computer application technology, in particular to a method for recognizing Uighur agricultural technical terms. Background technique [0002] At present, there is no automatic recognition method for Uyghur agricultural terms. Although the recognition methods for Uyghur terms other than agriculture use rule-based and statistical-based methods, or a combination of the two, but because this method is not sufficient Considering that Uyghur, as an agglutinative language, has the characteristics of language knowledge formed by rich language morphological changes, a large amount of annotated corpus is needed in the recognition process, and the recognition effect is overly dependent on the scale of the annotated corpus and the annotation results, resulting in the automatic recognition of domain terms. At the same time, due to the insufficient application of domain features based on language knowledge in the existing Uyghur...

AI Technical Summary

Problems solved by technology

[0002] At present, there is no automatic recognition method for Uyghur agricultural terms. Although the recognition methods for Uyghur terms other than agriculture use rule-based and statistical-based methods, or a combination of the two, but because this method is not sufficient Considering that Uyghur, as an agglutinative language, has the characteristics of language knowledge formed by rich language morphological changes, the recognition process requires the support of a large amount of labeled corpus, and the recognition effect is too dependent on the size of the labeled corpus and the labeling results, resulting in the automatic recognition of domain terms. At the same time, due to the insufficient application of domain features based on language knowledge in the existing Uyghur term recognition methods in other fields, the term extraction field is poorly targeted; there is also a lack of integration for term automatic The unified framework of the statistical features of recognition and language knowledge features, and the random use of various features lead to the problem of poor overall recognition effect. Therefore, this method is not suitable for automatic recognition of Uyghur agricultural terms.

Images