Institution name recognition processing method, intelligent terminal and storage medium

Abstract

The invention discloses an institution name recognition processing method, an intelligent terminal and a storage medium. The method comprises the steps that alphabet names with different part-of-speech in an institution name are predefined, and a word bank is created through collection and supplementation; preliminary division is performed on the institution name through an improved hidden Markovmodel, and the part-of-speech and geographic name division codes are marked; when it is detected that geographic names and trade names not existing in the word bank appear in the institution name, unregistered geographic names and trade names are recognized; a word segmentation result is further optimized through the configurable rule to meet custom demands of different users; and whether the unregistered words with the appearance frequency exceeding preset frequency are new industrial words is judged through statistical analysis, and if it is judged that the unregistered words are new industrial words, the unregistered words are automatically added into the word bank. Through the institution name recognition processing method, the intelligent terminal and the storage medium, effective andcorrect segmentation of the institution name and correct recognition of the part-of-speech are realized, and the recognition accuracy and recognition efficiency of a Chinese institution name are improved.

Description

technical field [0001] The invention relates to the technical field of Chinese information processing, in particular to an organization name identification processing method, an intelligent terminal and a storage medium. Background technique [0002] In recent years, with the development of natural language processing technology, more and more Chinese text processing tasks are automatically completed through programs. In text processing, a large number of institution names can be extracted; for example, when a user applies for a credit card, the bank needs to send the user The entered work unit is compared with the work unit of other information sources (such as the social security system), and due to reasons such as abbreviation, many Chinese organization names have inconsistent and non-standard expressions. For example, "ABCD Co., Ltd." may also be expressed as: "ABCD", "ABCD Company", "Guangdong Shenzhen ABCD Co., Ltd." and other names. Segment each element and identify ...

AI Technical Summary

Problems solved by technology

[0003] At present, the mature word segmentation system is mainly aimed at the news field and long texts. The processing of the organization name is mainly based on the overall recognition, and it cannot accurately segment the various elements in the organization name; especially the ubiquitous Business names generally do not exist in the thesaurus, and are called unregistered words; the current identification technology for unregistered words is generally the name of a person, place, or organization, and there is no automatic identification technology for unregistered business names; the identification of personal names, The surname is an obvious trigger point; the recognition of place names, the suffixes of place names such as "province" and "city" are obvious trigger points; and the composition of "business name" is ever-changing, and there is no obvious trigger point, which increases the difficulty of business name recognition. difficulty

[0004] That is to say, in the existing technology, the accuracy and efficiency of subdividing and segmenting Chinese organization names are not high. Many business names and place names are not accurately identified, and the elements in the organization name cannot be accurately segmented, which cannot be timely and effective. Identify new industry terms

Images