Unlock instant, AI-driven research and patent intelligence for your innovation.

Knowledge graph embedded multi-hop question and answer method,

A knowledge map and question technology, applied in the field of knowledge map question answering, can solve problems such as the inconsistency between embedding and entity embedding space, and the lack of embedding of natural language question word vectors, so as to achieve the effect of improving accuracy and accurate relationship representation

Pending Publication Date: 2022-02-11
GUILIN UNIV OF ELECTRONIC TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] What the present invention aims to solve is the problem that the question embedding of multi-hop question and answer is inconsistent with the entity embedding space and the word vector embedding of natural language questions lacks global information. It provides a multi-hop question and answer method for knowledge graph embedding

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
  • Knowledge graph embedded multi-hop question and answer method,
  • Knowledge graph embedded multi-hop question and answer method,
  • Knowledge graph embedded multi-hop question and answer method,

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0028] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with specific examples.

[0029] A knowledge graph embedding multi-hop question answering method, such as figure 1 and 2 As shown, its specific steps are as follows:

[0030] Step 1. Generate entity vector dictionary e and relationship vector dictionary r from the known knowledge map in the form of triples.

[0031] Step 1.1, the known knowledge map is stored as T=(e s ,r,e o ), where e s Represents the head entity, r represents the relationship, and e oIndicates the tail entity, such as (Yao Ming, wife, Ye Li), "Yao Ming" is the head entity, "wife" is the relationship, and "Ye Li" is the tail entity.

[0032] Step 1.2: Encode each entity and each relationship in the knowledge graph separately to obtain an encoded triplet.

[0033] Step 1.3: Use the ComplEx embedding model to train each enc...

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 knowledge graph embedded multi-hop question and answer method. Through the thought of a translation model, a vector space where a question q is embedded serves as a source sequence to be translated into a vector space where an entity is embedded, and the problem that the vector space where the question q is embedded is inconsistent with the vector space where the entity is embedded is solved. Through weighted calculation of a Tf-Idf weight coefficient, the problem that a word vector embedded in a word vector problem embedded in a problem q cannot consider global information of all question data in data is solved, so that prediction of answers is more reasonable. Therefore, the accuracy of multi-hop question and answer based on the knowledge graph is improved.

Description

technical field [0001] The invention relates to the technical field of knowledge graph question answering, in particular to a knowledge graph embedded multi-hop question answering method. Background technique [0002] The knowledge graph is a structured knowledge base containing triples (head entity, relationship, tail entity). Its essence is a directed graph with labels. The nodes in the graph represent entities and the edges represent relationships. Large-scale knowledge graphs include Freebase built in 2008, Wikidata built by Google in 2013, and DBPedia extracted from Wikipedia entries given by Lehmann et al. in 2015. Question Answering over Knowledge Graph (KGQA) is a research field that applies knowledge graph information, that is, given a natural language question and a knowledge graph, by analyzing the information contained in the question and knowledge graph, the knowledge graph question answering system tries to give come up with the correct answer. [0003] Early...

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): G06F16/332G06F16/36
CPCG06F16/3329G06F16/367
Inventor 李凤英陈明东董荣胜
Owner GUILIN UNIV OF ELECTRONIC TECH