Computer Science · Information Systems
Un système de recommandation hybride pour le parcours académique des étudiants, basé sur le RàPC et les ontologies
See on the mapAbstract
Academic advising at most high schools is limited in its ability to help students in identifying appropriate academic pathways. For example, choosing a university and a university major is a challenging task rife with anxiety that gets students confused. Therefore, students at high school need assistance to match their interests with the available universities and university majors. This thesis presents a novel hybrid recommender system (RS) approach for students’ academic advising supported by Case-based reasoning (CBR) and ontology. This Knowledge-based (KB) hybrid RS enables high school students to submit their demographic data, courses’ ratings, preferences, interests, and other knowledge in order to get personalized recommendations. These recommendations consist of three essential suggestions namely the university/college, university major, and career choice that match the student interests and requirements. The four major contributions of our research are: (1) Proposing a novel hybridization approach that incorporates four-core technologies namely the KB recommender system, Collaborative Filtering (CF), CBR, and ontology; (2) Overcoming the limitations and problems of the basic recommender systems’ techniques; (3) Designing the ontology that represents the high school student’s profile and interests, higher education and career domains; (4) Generating personalized recommendations based on the case-base and ontology; (5) Treating through our hybrid RS high dimensional datasets that contain heterogeneous data types. A prototype of this hybrid RS named (COHRS) is developed and presented. The experimental assessments and comparisons with the existing traditional recommender systems’ approaches, strongly demonstrate the high performance and accuracy of our hybrid RS that is supported by CBR and ontology.