Title: Social network and user context based recommender system architecture

Abstract:

In this seminar we will present a highly personalized recommender system where personalization data is acquired dynamically from user’s social networks. We also consider user’s current and historical context to improve the likelihood of acceptance of recommendations from user’s perspective. Personalized recommender systems are required to help people cope with the information glut in recommender systems. Our social networks activities already contains necessary likes, dislikes, interest information which can used for personalization of recommendation system. Our system architecture introduces a plug-and-play architecture which is easily adaptable to any ecommerce website. We also have described possible business models on our architecture which introduces direct and indirect business models. Such, business model can be considered a new dimension of user trust and commercial profit which is a novel approach.