Middle school experiment-oriented knowledge graph dynamic knowledge completion method

Abstract

The invention discloses a middle school experiment-oriented knowledge graph dynamic knowledge completion method, which comprises the following steps of: specifying a node propagation mode based on a graph neural network structure, completing the embedding of a knowledge graph, and solving the embedding problem of entities outside the knowledge graph through the information of neighbor nodes; obtaining a triad distance score based on a translation model TransH of representation learning, and preliminarily classifying triads; calculating a middle school experiment knowledge triple confidence coefficient; and completing triple classification of middle school experiment knowledge. According to the method, the problem of embedding entities outside the knowledge graph is solved through representation of nodes in the middle school experiment knowledge graph learned by the graph neural network, and the generalization ability of the knowledge completion method for new entities is improved; and further classification is carried out by adopting the triple confidence of the most similar relationship of the comprehensive semantics, and the semantic information of the middle school experiment is fully utilized, so that the accuracy of triple classification is further improved.

Description

technical field [0001] The invention relates to the technical field of knowledge graph completion, in particular to a method for dynamic knowledge completion of knowledge graphs oriented to middle school experiments. Background technique [0002] The knowledge graph is a new concept proposed by Google in 2012. It represents knowledge in the form of triples and is a knowledge base displayed in the form of a graph. Knowledge graphs are divided into general knowledge graphs and domain knowledge graphs. The latter brings together professional knowledge in a field and has high requirements for professionalism and accuracy. It can perform knowledge reasoning and assist decision-making in specific fields. Secondary school knowledge is knowledge that has been sorted out and formed into a system. This knowledge system has been verified and updated by several generations, and it is authoritative and correct. Middle school experimental knowledge is a part of middle school knowledge, a...

AI Technical Summary

Problems solved by technology

The task of the static knowledge completion method is to predict the missing relationship in the knowledge map, but it does not have the generalization ability, and it cannot effectively represent the entities that have not appeared in the knowledge map. Usually, all the knowledge maps including the knowledge subgraph need to be reconstructed. Training to complete the task will take a lot of time

Images