Knowledge graph question answering method, system and terminal based on entity relationship disambiguation

Abstract

The present invention relates to the field of information technology, and relates to a knowledge map question answering method, system and terminal based on entity relationship disambiguation; the method includes acquiring question text input by a user terminal, and identifying entity mentions in the question text Words, attribute mentions, and specific relationships; link entity mentions and attribute mentions into a pre-built knowledge graph, and build a semantic hypergraph based on the second-degree relationship subgraphs of each entity linked to; Utilize the multi-granularity context features of each entity included in the semantic hypergraph, and use the extreme gradient lifting algorithm to perform binary classification linear regression to jointly disambiguate entities and relationships; the present invention uses the first-degree relationship subtree and the second-degree relationship The evaluation score of the subtree avoids the problem that the implicitly expressed information in the questions is omitted in the present invention. Through the multi-granularity feature, the disambiguation ability of entities and relationships is greatly improved, and the accuracy of the system to answer questions is greatly improved.

Description

technical field [0001] The present invention relates to the field of information technology, in particular to the sub-field of natural language processing, and in particular to a knowledge base question answering method, device and terminal based on semantic hypergraph joint disambiguation and evaluation. Background technique [0002] With the rapid development of the Internet, all kinds of knowledge are becoming richer and even more explosively growing; and knowledge graphs can store massive amounts of knowledge well and are an important direction in the field of artificial intelligence. How to make good use of knowledge graphs to satisfy people's needs The need for knowledge acquisition also needs to be paid more and more attention. Therefore, it is very important to be able to query the knowledge graph through natural language questions to obtain the answers to the questions. The purpose of Knowledge Base Q&A is to provide people with a powerful knowledge acquisition too...

AI Technical Summary

Problems solved by technology

[0006] However, most of the existing entity disambiguation methods use the information of the mentioned words and the entity itself for disambiguation, while ignoring the information contained in the knowledge graph and questions.

In terms of query relationship disambiguation, the existing technology is to generate all possible results through artificially designed templates and rules, and then evaluate and sort them, resulting in low efficiency.

Images