Classifying User Predilections using Naïve Bayes Classifier (NBC) and Jaccard Similarity for Service Recommender System in Big Data Applications

Abstract

Service recommender systems have been shown as valuable tools for providing appropriate recommendations to users. The main objective is to identify a system that will classify the user reviews using effective methods and provide personalized recommendations to the users. The proposed architecture will present the different ratings and rankings of services to different users by considering diverse users' preferences, and therefore it will meet users' personalized requirements. The data classification can be achieved through analysing the user review as positive or negative using Naive Bayes classifier (NBC) in large-scale datasets and Jaccard Similarity and MinHash used to compute the similarity and provide the recommendation to user.