Medical named entity recognition system and method based on ALBERT model fusion

Abstract

The invention relates to a medical named entity recognition system and method based on ALBERT model fusion, and the system comprises a data preprocessing module which carries out the word segmentation of a question of a user, entity labeling and dictionary construction; a text coding module used for converting a text into an understandable data type and a calculation unit; a model modeling module used for constructing a framework of a model according to the task; an entity extraction module used for performing entity extraction and classification on the information subjected to feature extraction through model construction operation; and a dictionary construction module used for constructing a named entity corpus dictionary of the medical record. The method comprises the following steps: performing data preprocessing on an electronic medical record text, performing word segmentation on the text by a Chinese word segmentation module, labeling the text by an entity labeling module, and removing some wrong and useless data; and performing data cleaning on the electronic medical record, so that the model training time cost is effectively reduced. The novel model fusion mode solves the problem of named entity recognition in the field of electronic medical records in the medical field, and has higher entity recognition accuracy.

Description

technical field [0001] The invention relates to a medical named entity recognition system based on ALBERT model fusion and a method thereof. Background technique [0002] At present, with the rapid development of artificial intelligence technology, it is urgent to use existing technology to truly solve the problems in the real life of human society in order to truly benefit from technology. Electronic medical records are generated in the process of clinical treatment, in which named entities and entity relationships reflect the patient's health status, and contain a large amount of medical knowledge closely related to the patient's health status. Important extension. Judging whether a named entity is correctly identified includes two aspects: whether the boundary of the entity is correct, and whether the type of the entity is marked correctly. The main error types include that the text is correct, and the type may be wrong; conversely, the text boundary is wrong, but the m...

AI Technical Summary

Problems solved by technology

Compared with English, the task of Chinese named entity recognition is more complicated, and it is more difficult due to factors such as word segmentation; and the current named entities are carried out on the general corpus no matter in English or Chinese context, and for some professional fields, it is basically No application, e.g. medical field

Images