A Group-Centric Intelligence Recommendation System for Twitter

Abstract

The group-centric recommendation system develops logical collective results from the information given by Twitter users. Even though different input data formats have been used to represent user preferences, the input information mode is static. To avoid this shortcoming, this paper proposes a system which enables clients to give halfway or deficient inclination information at various occasions. Since this is an entangled issue, this paper explicitly centers around specific perspective (recommending movies) as the main endeavor. Accordingly, the re-analysis of variant input datasets, the maximum consensus mining problem, with the help of sentiment analysis, review analysis and rating analysis has merged singular proposals into clusters of suggestions under dynamic information mode suspicion. The outcome demonstrates that the proposed strategy is computationally productive and can adequately distinguish a general understanding among all clients.