PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
April 20, 2026Egyptian Informatics Journal1 citationsOpen Access

ALP-CRS: enhancing course recommendation in online education through data mining and graph analysis

View Full Paper
YGYanjun GaoXLXiaoyang LiuXGXinke Gao

Key Points

  • The aim is to improve course recommendations in online education using data mining and graph analysis techniques.
  • Developed a course recommendation system called ALP-CRS
  • Employed link prediction and graph clustering for accuracy
  • Used association rule mining to identify trends in course selection
  • Conducted three methodological stages: feature extraction, student graph creation, and recommendation generation
  • Achieved 82% precision rate and 24% recall rate in recommendations
  • Outperformed existing approaches by at least 4% in precision
  • Outperformed existing approaches by at least 10% in recall

Abstract

Recommender systems (RSs) are an increasingly common type of intelligent software that predicts the preferences of users. Online learning has become a ubiquitous and necessary means of lifelong learning, which provides flexibility in time and place. This paper introduces a course recommendation system (ALP-CRS) which is founded on APRIORi and Link Prediction. It employs link prediction and graph clustering techniques to enhance the accuracy of the recommendations, and employs association rule mining to detect common trends in the course selection. Three stages make up the methodology: outlining the guidelines for choosing courses, comparing student characteristics, and making suggestions. In the first stage, features are extracted from the course selection data and course selection rules are derived. In order to improve recommendations using link prediction, the next step comprises creating student graphs, grouping them, and filtering comparable students. During the last stage, the system offers recommendations on the filtered information, which gives students a more targeted and personalized experience of learning. According to the findings, the suggested approach was capable of providing suggestions with an 82% precision rate and a 24% recall rate, meaning that it outperformed the compared approaches by at least 4% in precision and 10% in recall.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Gao et al. (2026) studied this question.

synapsesocial.com/papers/69e5c2d003c2939914028d4dhttps://doi.org/10.1016/j.eij.2026.100955
Ask AI
Helpful
Bookmark
Share
View Full Paper