Adaptive Intelligent Recommendation System Based on Personality Mining and Knowledge Graph in An Online Learning Environment

Abstract

This article reviews the recent research work and sum up shortcomings in existing intelligence
recommendation systems. Then, we comprehensively examined the various elements of individual
characteristics and online teaching methods. Individual characteristics include more than 30 indicators in 6
aspects. There are more than 10 online teaching methods for intelligent selection. Next, we use data mining
technology to comprehensively mine students' personalized features and online learning time features;
define multiple node ontology in knowledge graph (KG) and construct a three-layer digital KG model;
design two-level intelligent matching, pruning and anti-transition pruning Algorithms based on personalized
learning space-time features and teaching methods; give the method of system adaptive dynamic update and
development based on environment perception and the overall functional framework of the system. Finally,
experimental results validate the effectiveness of our recommendation system. The value of our
achievements mainly includes: understanding students' characteristics comprehensively in intelligent
education research; cross-language application; further improving the recall rate and precision rate; realizing
customized intelligent recommendation of “one person, one teaching”; multi-faceted machine learning and
self-development of intelligent recommendation system.