ABSTRACT
One of the most intriguing studies in the world of education is custom online learning based on a recommender system. The majority of academics created a variety of recommender e-learning approaches that use recommendation methods in mining of educational data particularly for the determination of learning choices of students. But it doesn’t produce satisfactory outcomes. The suggested system designed an Enhanced Vector Space Model (ESVM) based recommendation for the Protus system in order to improve suggestion accuracy and reduce query processing time. The learning styles of students are initially derived from server blogs. Similarity computation and recommendations are carried out when preprocessing is finished. The system first creates a suggestion list using content-based filtering. Based on the adjusted cosine similarity of their content, the results are ranked. In order to accurately classify the most active participant in the group, we use a collaborative method. The experimental results prove that the designed EVSM scheme performs better than the old system in terms of query processing speed, MAE, and correctness.
