Map-Reduce Based Parallel Firefly Algorithm For Fast Recommendations

Abstract

The Recommendation System is a strong tool that aids the decision-making process across a variety of situations. Using the aid of aspects such as prior experiences of the user, their ratings, comparable interests, etc., we can acquire the most relevant results upon the application of various optimization techniques. A movie recommendation system is a useful tool/software that aids users in rapidly obtaining optimum results with comparable interests. Using the Firefly clustering technique, this study focuses on a movie recommendation system whose major goal is to propose movies of comparable interest to the active user. Although much study has been done in the topic of recommendation systems, there are still several issues with producing suitable results. To address these issues, we suggested a strategy that uses a meta-heuristic approach to get optimal outcomes. Instead of utilising K-means, fuzzy C-means, and other algorithms, we present the Firefly clustering method in this research to provide the best optimum outcomes in recommendation systems. For performance analysis, many measurements such as t-value, RMSE, SD, and MAE are utilised.