ENHANCING MOVIE RECOMMENDATION USING A MODIFIED ALTERNATING LEAST SQUARES ALGORITHM
Akilu Saadatu Sabo, Iliyasu Badariyya Sani, 3Shehu Nafisa Abubakar, Muhmmmad, Aishatu Abubakar, Muhammad, Amatullah Sanusi, Mohammed, Maikolo
Abstract
ABSTRACT Recommender systems are software techniques and tools that provide suggestions of items to be used by users, the suggestions are aimed at supporting users in various decision making process. The study tried to enhance a movie recommender system using a Modified Alternating Least Squares Algorithm to improve recommendations. An experiment was conducted to analyze the performance of the proposed design. The result shows that the algorithm effectively improves the movie recommender system. To validate system, we ran extensive experiments using Apache Spark with Movielens1M dataset. We compare the system with the other methodologies. The performance metrics used in the experiments are Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Precision and Recall. The results of the experiments show that the proposed Modified Alternating Least Squares (MALS) algorithm reduces the RMSE, MAE values compared to the existing recent ALS-based Algorithms, thereby improving the recommendation accuracy of the recommender system. As such, MALS is recommended to be used in training collaborative filtering recommendation systems, so as to result in generating best recommendations.