Intelligent error correction method and device for proper nouns, equipment and storage medium

Abstract

The invention relates to the technical field of big data. The invention discloses an intelligent error correction method for proper nouns. The method comprises the steps of obtaining a proper noun tobe corrected, performing word segmentation processing on the proper noun to be corrected to obtain a plurality of word segmentation fragments of a text to be corrected, outputting the word segmentation fragments in a pinyin format, taking pinyin of each word segmentation fragment as a keyword, and retrieving candidate words corresponding to the word segmentation fragments from a preset homophone dictionary to obtain a retrieval result; if the retrieval result is not empty, determining each retrieval candidate word based on the retrieval result, respectively calculating scores of the retrievalcandidate words, sorting the scores, and outputting a sorting result; and based on the sorting result, taking the candidate word with the highest score as a replacement item to replace the corresponding word segmentation segment. The invention further discloses an intelligent error correction device and equipment for the proper nouns and a computer readable storage medium. According to the invention, a more accurate intelligent error correction service of the proper noun is provided for the user, and the error correction efficiency is improved.

Description

technical field

[0001] The present invention relates to the field of big data technology, in particular to an intelligent error correction method, device, equipment and computer-readable storage medium for proper nouns. Background technique

[0002] In recent years, with the continuous development of the economy and society, in the vertical field, when customers consult questions, they often aim at a specific professional direction. These similar questions often include some special terminology in this field. Users often make typos or some typos appear in these proper nouns due to language conversion, which will make it difficult for subsequent modules to accurately judge the user's true expression.

[0003] In the current input method or language recognition, to correct errors in proper nouns caused by users’ typos or language conversion, most of them rely on sequence labeling models, such as ner, lstm+crf, etc., but because of this method A large number of aligned corpora...

Images