Named entity recognition method and system for medical health knowledge map

Abstract

The invention discloses a naming entity recognition method and a naming entity recognition system for a medical health knowledge map. The naming entity recognition method comprises the following steps: tagging a training corpus and generating a word vector according to the tagged training corpus; labeling a training corpus according to the tagged training corpus; training a named entity recognition model using the word vector, wherein the named entity recognition model is based on an iterative void convolution neural network ID-CNN and CRF; the recognition results of the sentences to be testedare obtained through the trained named entity recognition model. By using the ID-CNN and CRF, the named entity recognition model of CNN and CRF can solve the problem of named entity recognition in the construction of medical health knowledge map in the health domain, and some specific knowledge entities can be automatically recognized for the text-based corpora in the health management domain, which can be used in the medical health knowledge map.

Description

technical field [0001] The present invention relates to a method and system for automatically identifying health knowledge entities from medical and health texts, in particular, to a named entity recognition method and system for medical and health knowledge graphs. Background technique [0002] Knowledge graph is a visual display form using network graph to associate knowledge between multiple disciplines in different fields, aiming to describe concepts, entities, events and the relationship between them in the objective world. The health management field is much more dependent on knowledge than most text information processing applications, which makes establishing a richer and more accurate medical and health knowledge map an important research content. [0003] Different from the traditional expert knowledge base, the medical and health knowledge map needs to automatically extract the target knowledge from large-scale multi-modal health management information, and establ...

AI Technical Summary

Problems solved by technology

[0005] In view of the above-mentioned problems in related technologies, the present invention proposes a method and system for identifying named entities in the field of medical and health knowledge by using ID-CNN and CRF models, which can solve the problem of named entity recognition in the construction of medical and health knowledge graphs

Images