MOVBOK: A Personalized Social Network Based Cross Domain Recommender System
Sharu Vinayak, Richa Sharma, Rahul Kumar Singh · Indian Journal of Science and Technology · 2016
Objective: We propose a novel idea for resolving research issues like cross domain recommendations and recommendations using social networks in the emerging research field recommender systems. Methods/Analysis: According to this idea user will be recommended with the list of books that belong to the genre that is most liked by the user in terms of movies. Findings: Here we will collect user’s tastes in movies from his social network profile and extract out the most liked genre by him and using an appropriate collaborative filtering algorithm will recommend him with the books that may interest him. Improvement: The proposed idea is expected to resolve research problems like cold start problem and sparsity. Our proposed methodology gives more competent results than the traditional. Keywords: Cross Domain Recommendations, F1 Score, Precision, Recall, Recommender Systems, Root Mean Square Error