Knowledge graph representation learning method through combination of entity hierarchy category

Abstract

The invention relates to a knowledge graph representation learning method through combination of the entity hierarchy category. The knowledge graph representation learning method comprises the steps that the trituple relationship of a knowledge graph and the hierarchical structure category information of the entity are acquired; the category mapping matrix of the entity under the preset trituple is constructed according to the hierarchical structure category information of the entity; an energy equation is constructed according to the entity vector and the relationship vector of the trituple relationship and the category mapping matrix; and a margin-based evaluation function is constructed according to the energy equation, and representation of the entity vector, the relationship vector and the category mapping matrix is learnt by minimizing the evaluation function. According to the knowledge graph representation learning method through combination of the entity hierarchy category, the representation learning effect can be enhanced by fully utilizing the category information of the entity having the hierarchical structure, the higher accuracy can be obtained in the task of knowledge graph completion and trituple relationship classification and effect enhancement is especially prominent in the low frequency trituple relationship having long-tailed distribution so that the method has great practicality.

Description

technical field [0001] The present invention relates to the technical field of natural language processing, in particular to a knowledge map representation learning method combined with entity hierarchical categories. Background technique [0002] We are currently in the era of information explosion. With the rapid development of society, massive amounts of knowledge and information are generated every day. This information is usually generated and stored in unstructured forms such as text or pictures, while applications such as information retrieval and question answering systems require more accurate structured information. With the increasing demand of users for information screening and sorting, how to mine valuable information from massive data has become a difficult problem. Therefore, the knowledge map came into being. [0003] The knowledge graph aims to build a database of structured information, expressing concrete things in the world (such as proper nouns such a...

AI Technical Summary

Problems solved by technology

[0007] The technical problem to be solved by the present invention is to provide a knowledge map representation learning method combined with entity hierarchy categories, which solves the problem of not making full use of entity hierarchy category information in the prior art, and improves the representation performance of knowledge maps

Images