Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Chinese medical text entity relationship joint extraction method based on conversation attention mechanism

An entity relationship and attention technology, applied in neural learning methods, text database query, unstructured text data retrieval, etc., can solve problems such as information waste, performance degradation of relationship extraction models, and the inability of traditional models to provide better solutions. , to achieve the effect of strengthening the connection and improving the accuracy

Pending Publication Date: 2022-07-15
ZHEJIANG UNIVERSITY OF SCIENCE AND TECHNOLOGY
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The traditional pipeline model needs to use real entity labels for training, while the output of the entity recognition model is used in the relationship extraction stage. The difference in distribution between the two will lead to a decrease in the performance of the relationship extraction model.
In fact, there is an implicit relationship between the entity type and the relationship type, and the pipeline method does not take advantage of such a relationship
Moreover, the pipeline method performs relationship extraction for each entity pair, resulting in a lot of waste of information
Moreover, the traditional model cannot provide a better solution to the problem of overlapping entity relationships.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Chinese medical text entity relationship joint extraction method based on conversation attention mechanism
  • Chinese medical text entity relationship joint extraction method based on conversation attention mechanism
  • Chinese medical text entity relationship joint extraction method based on conversation attention mechanism

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0044] The present invention is further analyzed below in conjunction with specific embodiments.

[0045] The technology provided by the present invention is a method for jointly extracting entity relationships in Chinese medical texts based on a conversational attention mechanism, comprising the following steps:

[0046] Step 1. Input the sentence into the RoBERTa layer to fully extract the sentence features and mine the association between words:

[0047] The overall architecture of this model is roughly divided into two parts: the first part is the RoBERTa layer; the second part mainly includes the CLN layer and the THA layer, which predicts the corresponding object entity according to each relationship corresponding to the main entity. figure 1 The general architecture of this model is shown. The sentences are input into the RoBERTa layer to fully extract sentence features and mine word-to-word associations. Extracting the head entity and the tail entity is performed in th...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a Chinese medical text entity relationship joint extraction method based on a conversation attention mechanism. According to the method, the thought of carrying out feature fusion on the CLN layer and the position information is provided, and a Talking head attention mechanism is introduced, so that conversation type interaction is carried out among all relationships. The relation between the entity type and the relation type is enhanced, and the accuracy of the model is greatly improved.

Description

technical field [0001] The invention belongs to the technical field of computer applications, and relates to a method for joint extraction of entity relations in Chinese medical texts based on a conversation attention mechanism. Background technique [0002] The medical knowledge graph is constructed according to the knowledge in the medical field. It aims to organize the knowledge in the medical text systematically by establishing the association relationship between medical entities, so as to provide convenience for downstream data search, mining and analysis. The medical field has a large amount of textual information, but how to extract the required medical knowledge from these medical texts to construct a knowledge graph has become a hot research topic. [0003] The construction of knowledge graph is inseparable from information extraction (IE). The research difficulties in information extraction (IE) are named entity recognition (NER) and entity relation extraction (RE...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/35G06K9/62G06F16/33G06F40/295G06F16/36G06N5/02G06N3/04G06N3/08
CPCG06F16/35G06F16/3347G06F40/295G06F16/367G06N5/02G06N3/08G06N3/047G06N3/048G06F18/2415G06F18/241G06F18/253
Inventor 黄杰罗之宇张蕾万健史斌彬张丽娟
Owner ZHEJIANG UNIVERSITY OF SCIENCE AND TECHNOLOGY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com