Relation prediction method combining logic rules and fragmented knowledge

Abstract

The invention relates to a relation prediction method combining logic rules and fragmented knowledge. The method comprises the following steps: firstly, performing unified modeling on a fact triple and a logic rule, and embedding hidden semantic information into a knowledge representation-based relationship reasoning model; secondly, in combination with fragmented knowledge, performing continuousiterative updating, so that a knowledge base becomes more complete. According to the method, unified modeling is carried out on the fact triad and the logic rule, hidden semantic information is embedded into the relation reasoning model based on knowledge representation, and more accurate prediction is achieved.

Description

technical field [0001] The invention relates to a relationship prediction method combining logic rules and fragmented knowledge. Background technique [0002] In the field of relational reasoning, the relational reasoning model represented by TransE [1] has become a research hotspot in recent years because of its simplicity, efficiency, and good predictive performance. The TransE model directly models the fact triples (h, r, t) in the knowledge base. The basic idea is to map the entities and relationships in the knowledge base to a low-dimensional continuous vector space, thereby simplifying the knowledge base. related calculations. Although the basic representation learning model is simple and efficient, it only considers the direct fact triples (h, r, t) in the knowledge base and ignores the hidden semantic information in the knowledge base, resulting in limited inference accuracy. Some recent work utilizes adding external data, such as entity types, textual descriptions...

AI Technical Summary

Problems solved by technology

However, most existing knowledge representation learning methods only use fact triples to perform embedding, ignoring some hidden semantic information in the knowledge network, which not only makes the learned vectors unable to accurately express the semantic relationship in the original knowledge base, but also Cannot make full use of the value brought by fragmented knowledge

Images