A Comprehensive Survey Using Novel Content-Based Filtering Technique to Improve Effectiveness on Social Media Datasets for Group Recommendation System Based on External Social Trust Networks

Abstract

Aim:To build a Movie Recommendation system for Netflix to improve the accuracy percentage using Content-Based Filtering technique to analyze the prediction rate of movies. Materials and Methods: Two different algorithms are used Content-Based filtering algorithm and External Social Trust Networks to achieve maximum accuracy with a content based filtering with a sample size = 6 and external social trust network with a sample size= 6 was iterated 10 times for accurate outcome. Pre test is calculated by using G-Power tool value of 80% and confidence interval is 95% mean and standard deviation. Result and Discussions: These results proved that Recommendation System for Netflix using Novel Content-Based filtering algorithm gives better accuracy (0.94) in comparison with External Social Trust Networks(0.76). Conclusion: Content-Based Filtering algorithm helps to build a recommendation system gives effective results with the better accuracy rate of percentage.