Deep Learning Framework for Student Course Recommender System for Higher Educational Institutions

With the advancement in technology like Artificial Intelligence (AI) and Machine Learning (ML) there are many sub fields which are emerged from the core areas. Deep Learning (DL) is one such field which is gaining vast importance and is often mentioned as branch of ML and AI. In order to quickly and efficiently find the relevant information on the internet based on our past browsing history, recommender systems are playing a major role. Many such systems has been proposed in recent works. Recommender system has been implemented in major E-commerce sites like Amazon, Flipkart, Netflix, YouTube and many other giants have introduced AI based recommendation systems. Most of these systems uses the algorithms which can efficiently collect user behavioral data, filter it and recommends the user various choices based on the previous data. These systems often need to develop their own framework and algorithms which can solve the purpose. Students are often confused to opt for courses out of various options available in today's digital world. Based on their skill, interest the system should recommend a course. A general framework where it can handle data diversity, various dimensions and stimulation types is needed. Deep learning based generalized framework can solve such type of issues. In this paper we propose such generalized and effective Deep learning framework for a student course recommendation system for students who wish to pursue a course in various Higher Educational Institutions. The recommender system can be useful for the students who wish to opt a particular course based on their past history of interest and data collected from the logs. In this paper we use the Intel OpenVino toolkit for the deployment of the recommender system. The recommender system which is based on deep learning will provide the student with the exact set of modules he need to choose based on his skill set. The result of a student recommender system using deep learning can vary depending on several factors, including the quality of the data, the effectiveness of the model, and the specific goals of the recommender system. In this paper we focused on better performance, personalized learning paths and enhanced student satisfaction.