Abstract
The very idea of Felicitous Area Recommender Systems has been actively researched during the last several decades, and lots of sophisticated models have been created. This paper introduces an integrated Academic Recommender System (ARS) that will help in capturing the multidimensional association among students and academic institutions through academic profiles, academic feedback, and socio-economic parameters. The proposed framework models such interactions as a heterogeneous graph and uses Graph Attention Networks (GAT) to solve important academic decision-making problems, such as programme selection (BE/MBBS/BSc, etc.), elective course recommendation, and personalized placement-training. Compared to the traditional methods of recommendations, our approach proves to be more accurate and predictive.