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

Clinical knowledge graph link prediction method and system based on relational graph attention network

A prediction method and attention technology, applied in medical data mining, special data processing applications, unstructured text data retrieval, etc., can solve the problem of low data quality of electronic medical record data, lack of relationship between entities, and inability to directly observe and other issues to achieve the effect of ensuring accuracy and comprehensiveness, fast training speed and few parameters

Pending Publication Date: 2021-11-30
SHANDONG NORMAL UNIV
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The inventors found that since the patient's medical record data is constantly growing, even the medical record data containing hundreds of millions of information contains incomplete information, and there are some implicit relationships between entities that cannot be directly observed
In addition, the actual electronic medical record data generally has the characteristics of low data quality, which makes there may be some missing relationships between entities in the clinical field data

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
  • Clinical knowledge graph link prediction method and system based on relational graph attention network
  • Clinical knowledge graph link prediction method and system based on relational graph attention network
  • Clinical knowledge graph link prediction method and system based on relational graph attention network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0034] The object of this embodiment is to provide a clinical knowledge chart link prediction method based on a relational graph pointing network.

[0035] A clinical knowledge chart link prediction method based on relational graph attention to network, including:

[0036] Get a clinical field knowledge map, and the entities and relationship information in the knowledge map are given a matrix representation;

[0037] The matrix of the entity and the relationship represents the input of the pre-training relationship diagram focused network model for learning, obtaining the exact vector representation of the entity and the relationship, and uses a two-line model and the embedding method of the translation model to form a triplet vector. Express;

[0038] Based on the convkB model and a predetermined rating function, the obtained ternary set vector represents the score, determines whether or not there is a link relationship between the patient's medical record entity based on the sco...

Embodiment 2

[0092] The object of this embodiment is to provide a clinical knowledge chart link prediction system based on a map-based network.

[0093] A clinical knowledge chart link prediction system based on a graphic branch network, including:

[0094] The data acquisition unit is used to obtain a clinical field knowledge map, and the entities and relationship information in the knowledge map are represented by matrix;

[0095] The encoding unit is configured to learn the entity and relationship matrix representation of the input pre-training relationship diagram, and obtain an accurate vector representation of entities and relationships, and adopt embedded embedded based on bilinear model and translation model. The method forms a triplet vector representation;

[0096] The decoding unit is used to score the obtained ternary set vector representation based on the CONVKB model and the predetermined score function, determine whether there is a link relationship between the patient's medical...

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 provides a clinical knowledge graph link prediction method and system based on a relational graph attention network, and the scheme achieves the link prediction of a medical record knowledge graph of a clinical medical patient based on the architecture of an encoder and a decoder. According to the scheme, a relational graph attention network is adopted, patient medical record information in a knowledge graph is coded, and vector representation is carried out on a triple in a more effective mode; meanwhile, a novel entity and relation embedded representation is provided; and finally, by using a ConvKB model, a scoring function is defined, the possibility of effective triples is judged, and according to the ranking of the triples, whether a link relationship exists between patient medical record entities is inferred, so that a link prediction task is completed.

Description

Technical field [0001] The present disclosure belongs to the field of knowledge chart link prediction technology, and more particularly to a clinical knowledge chart link prediction method and system based on a relational graph and focus network. Background technique [0002] The statement of this section is merely the background technology information related to the present disclosure, which is not necessarily constituted in prior art. [0003] The knowledge map consists of two parts of the node and the edge, and the node is used to represent the entity in the figure, and the relationship between the entity is used to indicate the relationship between the entities, and organize it into a three-component group by using the node and the side of the scattered knowledge information. Presented in structured forms, constitute a powerful semantic information network, thereby providing a better use of this information to solve some of the practical problems. Knowledge maps have a wide r...

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): G06F16/36G16H50/70
CPCG06F16/367G16H50/70
Inventor 骆超曹战月
Owner SHANDONG NORMAL UNIV
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