A knowledge graph question answering method and system based on graph neural network embedding matching

Abstract

The invention discloses a knowledge map question answering method based on graph neural network embedding and matching, which includes: acquiring a question from a user, processing the question with a named entity recognition tool to obtain entities in the question, and using The syntactic analysis tool processes the question sentence to obtain the query graph and the subject word corresponding to the question sentence; uses the entity synonyms dictionary to perform entity link processing on the obtained subject term to obtain the subject term in the knowledge map, and integrates the subject term in the knowledge map The subject words are entered into the knowledge map for retrieval to obtain the topic map, and the obtained topic map and the obtained query map are input into the trained graph embedding matching model to obtain the answers to the questions. The present invention can solve the technical problem that the template of the existing semantic analysis method cannot be fully applied to all natural language questions, and the fuzzy search ability for similar semantics is weak, and many unnecessary query sentences will be generated, resulting in low query efficiency technical problems.

Description

technical field [0001] The invention belongs to the technical field of artificial intelligence, and more specifically relates to a knowledge graph question answering method and system based on graph neural network embedding and matching. Background technique [0002] Integrating human knowledge is one of the research directions of artificial intelligence. Knowledge representation and reasoning, inspired by human problem solving, is to represent knowledge for intelligent systems to gain the ability to solve complex tasks. In recent years, knowledge graphs, as a form of structured human knowledge, have attracted great attention from academia and industry. A knowledge graph is composed of some interconnected entities and their attributes. In other words, the knowledge map is composed of pieces of knowledge, and each piece of knowledge is represented as a triplet, entity, relationship, entity (Entity-Relation-Entity). Entities can be both real-world objects and abstract conce...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the above deficiencies or improvement needs of the prior art, the present invention provides a knowledge graph question answering method and system based on graph neural network embedding and matching. The technical problem of language questions, the existing semantic analysis method is weak for the fuzzy search with similar semantics, and will generate many unnecessary query sentences, which will lead to the technical problem of low query efficiency, and the existing information retrieval method is due to the simple Analysis from the perspective of one-way sequence structure leads to the inability to learn complete semantic features, which in turn leads to technical problems of low query accuracy

Images