At a time of international trade friction, college teaching, especially English teaching, needs to shoulder the responsibility of enhancing students’ multicultural comprehension ability, so as to send more tolerant talents to the society in the future. This study uses the LDA model to extract the multicultural understanding themes and other contents in students’ English learning data to complete the students’ ability portrait. A knowledge graph is constructed to deeply explore the association between learning data and comprehension ability. Students are classified by K-means clustering algorithm, and personalized enhancement resources are provided for different types of students by combining the intelligent recommendation of knowledge graph resources. Students are best classified when they are clustered into 4 categories according to their major learning achievements. Knowledge mapping provides 5 paths for different types of students, which can meet the competence enhancement needs of different types of students.
