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

Method and device for noise detection and knowledge completion of knowledge graph

A knowledge map and noise technology, applied in the field of knowledge map data processing, can solve the problems of low robustness of the model and no consideration of auxiliary information, etc., to achieve the effect of improving versatility, improving effect, and reliable judgment

Pending Publication Date: 2021-04-30
NAT UNIV OF DEFENSE TECH
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the three-stage reliability estimator only considers the information of the internal structure and does not consider the auxiliary information, resulting in low robustness of the model

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
  • Method and device for noise detection and knowledge completion of knowledge graph
  • Method and device for noise detection and knowledge completion of knowledge graph
  • Method and device for noise detection and knowledge completion of knowledge graph

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0059] Such as figure 1 As shown, a method for noise detection and knowledge completion of knowledge graphs, including the following steps:

[0060] Step 1, obtain the data of the knowledge graph containing noise;

[0061] Step 2, projecting entities and relations to a low-dimensional space based on the translation framework;

[0062] Step 3, introducing entity type hierarchy information and relationship path information;

[0063] Step 4, calculating the matching degree of entities and relations in triples;

[0064] Step 5, calculating the credibility of the matching degree;

[0065] Step 6, calculating the triplet score based on the matching degree and credibility;

[0066] The model frame DSKRL of the present invention is composed of a triplet difference estimator and a triplet support estimator. The degree of difference and the degree of support describe the degree of matching of triplets and the credibility of the degree of matching, which can be measured by structura...

Embodiment 2

[0122] This embodiment provides a device for noise detection and knowledge completion of knowledge graphs, including one or more processors;

[0123] storage means for storing one or more programs,

[0124] When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in Embodiment 1.

[0125] The beneficial effects of the present invention are as follows:

[0126] (1) A basic framework for knowledge graph noise detection and knowledge completion is designed that integrates structural information, entity type hierarchy information, and relational path information. Entity type level information, relationship path information and structural information complement each other. This basic framework can greatly improve the effect of knowledge graph noise detection and knowledge completion, and then have a positive impact on downstream tasks and applications.

[0127] (2) Fewer hyperparameters are used, which ...

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 method and device for noise detection and knowledge generation completion of a knowledge graph. The method comprises the following steps: acquiring data of the knowledge graph containing noise; projecting entities and relationships to a low-dimensional space based on a translation framework; introducing entity type hierarchical information and relationship path information; calculating the matching degree of entities and relationships in the triple; calculating the credibility of the matching degree; and calculating a triple score by integrating the matching degree and the credibility. A knowledge graph noise detection and knowledge completion basic framework integrating structure information, entity type hierarchical information and relationship path information is designed; whether a triple is true or not is judged in two steps by utilizing entity type hierarchical information and relation path information, namely auxiliary information of the knowledge graph, so that noise existing in the knowledge graph is detected, and better knowledge representation is generated.

Description

technical field [0001] The invention belongs to the field of knowledge graph data processing, and in particular relates to a method and equipment for noise detection and knowledge completion of knowledge graphs. Background technique [0002] Knowledge graph (Knowledge Graph, KG) has been widely used in the real world, for example, knowledge-driven artificial intelligence and question answering systems. A typical knowledge graph usually contains a large number of triples to store knowledge, in the form of (head entity, relation, tail entity), which can be abbreviated as (h, r, t). In recent years, there are many widely used knowledge graphs in the real world, such as knowledge graphs in different fields such as Freebase, WikiData, WordNet, etc. Although we are still unable to achieve full coverage of knowledge in the real world, we must keep the knowledge graph updated to reflect changes in the real world, which is crucial for knowledge-driven applications. Traditional meth...

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
IPC IPC(8): G06F16/36G06F16/28
CPCG06F16/288G06F16/367
Inventor 赵翔谭真邵天阳李硕豪郭得科肖卫东张军
Owner NAT UNIV OF DEFENSE TECH
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