Named entity recognition method for Chinese medical text

Abstract

The invention belongs to the technical field of medical text labeling, and particularly relates to a method for identifying named entities of Chinese medical texts. According to the method, the entities in the original medical text are automatically annotated by customizing a plurality of entity categories and constructing the medical term annotation dictionary according to the entity categories;on this basis, a multi-granularity feature fusion model is provided. The radicals of the Chinese characters are used as entity recognition and classification features to be applied to medical entity recognition tasks for the first time; features on three different granularities of words, characters and radicals of the characters in the medical texts are extracted, expressed and fused, and a modelis trained by utilizing an ID-CNN-CRF algorithm, so that the recognition work of medical entities in various medical texts is realized. The method has the advantages that the method can be applied tovarious medical texts such as electronic medical records and medical periodicals. Meanwhile, the problem that in the medical field, the length difference between different entities is large can be well solved, and the method has a good effect on recognition of the entities which are not logged in.

Description

technical field [0001] The invention belongs to the technical field of medical text labeling, and in particular relates to a method for named entity recognition of Chinese medical texts. Background technique [0002] The annotation of medical text is a hot issue in the application of natural language processing in the medical field. Medical texts mainly include medical journals and electronic medical records formed during the medical treatment process. Medical texts are considered to be the core data of medical information systems, so it is very important to use computer programs to automatically mine these knowledge from these texts. This technology mainly involves the use of program codes to apply natural language processing (NLP), information extraction (including entities and relationships, etc.) and other related technologies on medical texts for analysis and mining. [0003] Medical named entity recognition is one of the important tasks in the application of natural l...

AI Technical Summary

Problems solved by technology

[0006](2) There are often many different modifiers and qualifiers in the context of the entity to be recognized and extracted, which makes it difficult to divide and determine the boundary of the entity

[0008](4) There may be huge differences in the length of different entities. For some disease names and drug names, the length of entities may be very long, and some entities may even Contains more than 10 characters, and some entities only contain 2-3 characters, which obviously increases the difficulty of the labeling task

Images