Personalized learning of college English using knowledge graphs combined with user portraits
Abstract
In this study, user portraits were combined with knowledge graphs to develop a personalized recommendation algorithm for English exercises and applied it to the teaching of English. A case analysis was conducted using sophomore students from the School of Foreign Languages at Ningxia Medical University. The optimal parameters of the long short-term memory (LSTM) algorithm for classifying exercise knowledge points were tested first, and a knowledge graph of English exercises was then constructed. The students were divided into a control group and an experimental group. The control group received traditional multimedia teaching, while the experimental group utilized the personalized recommendation algorithm for
English exercises to facilitate their learning. English tests were conducted before and after a four-week teaching period. Moreover, a questionnaire was administered to the experimental group to gather feedback on the new teaching method. The results showed that the LSTM algorithm performed best in classifying exercise knowledge points when the number of nodes in the hidden layer was 128, and the activation function was sigmoid. The teaching mode, assisted by the personalized recommendation algorithm based on the user portrait and knowledge graph, effectively improved students’ English
scores and increased their interest in learning English.
Keywords: user portrait, knowledge graph, English, personalized recommendation
Cite As
N. Wang, "Personalized learning of college English using knowledge graphs combined with user portraits", Engineering Intelligent Systems, vol. 33 no. 3, pp. 339-343, 2025.