A Hybrid Approach to Social Context Based Scalable Recommender System

Abstract

Social networking sites produce an exponential amount of information and it is necessary to incorporate scalable recommender systems. In Existing Recommendation Techniques, Social Relationship Information has not been fully used. It is hard to find the contextual factors and integrate them into efficient recommendation framework. In addition to that, Prediction qualities are the most important challenges. This paper analyses this recommendation problem on the basis of psychology of the person and social relation among the persons, that exhibit two factors – User’s Preference and Relationship Influence. It proposes Hybrid Filtering Approach to incorporate user’s preference and relationship influence. This paper uses the priority based ranking algorithm to improve accuracy of social recommendation.