Research on Network Personalized Learning Recommendation Algorithm Based on Knowledge Representation

Since 2000, education reform has quietly risen. As one of the research hotspots of education reform, individualized learning has been valued by various countries. The emergence of the online learning space provides a huge technical environment support for personalized learning. However, how to recommend personalized learning content for learners from the numerous resources on the Internet has always been one of the difficulties in personalized online learning. Technology is a key element for achieving personalized online learning. At present, there are fewer systematic researches and practical applications that combine personalized learning with online learning space. In order to better realize the personalized learning of students and make the personalized learning develop to a higher level, this paper studies the personalized recommendation. The research of personalized recommendation system is an important breakthrough in the realization of information filtering. It is usually based on the user's historical interaction information to mine potential interests and preferences to push items that satisfy users. In this paper, the knowledge representation learning method is used to retain the semantic information of the user or the item itself, and the data is embedded in the low-dimensional vector space to calculate the semantic similarity to generate the recommendation result. This algorithm enriches the semantic information of users or items in traditional recommendation algorithms, effectively enhances the recommendation performance, and makes better recommendations for network personalized learning.