Recommender System for enrollment advice of elective courses for students of Electronic Engineering and Telecommunications Engineering from UNAD

Currently, due to the pressing need for educational quality and coverage in Colombia, the academic sector is looking for remote learning experiences based on E-Learning methodologies. The Universidad Nacional Abierta y a Distancia - UNAD is one of the institutions that has stood out in the country for providing quality E-Learning remote education. Aiming to improve the student’s experience in the elective course enrollment process, this paper presents the development of a recommender system based on collaborative filtering capable of advising the students which elective courses are more suitable for them based on their academic history. Initially, the description of the database used to train and test the system is presented; then, the preprocessing done to the data is explained; after that, the development of two recommender models is shown, one based on Non-negative Matrix Factorization (NNMF) and one based on Singular Value Decomposition (SVD) matrix factorization; the, the results and evaluation of the proposed models is presented; and lastly, the conclusions of this work. Based on the results, it was selected the recommender system based on SVD matrix factorization, this model was able to suggest elective courses for the students with an 80% similarity when comparing its predictions with the recommendations of a human advisor.