A scalable solution for personalized recommendations in large-scale social networks
Christos Sardianos, Iraklis Varlamis · 2014
The modern trend in many Web 2.0 applications is that users can interact with the applications in terms of social activity, for example they can express their trust for another user or another user's review etc. Bearing that in mind, creating recommendations for users could go one step further by reclaiming the social network information they share into these applications. This information can be exploited for creating better recommendations but also for optimizing the execution of existing algorithms in large scale datasets. In this paper, we introduce a scalable model for generating personalized user recommendations, which first uses the social information and divides the social graph into subgraphs and then applies Collaborative Filtering on the preferences information related to each subgraph. The results of our experiments show that our proposed model can perform faster than the traditional techniques while producing high level personalized recommendations. Moreover, it allows the parallel execution of the recommendation algorithm in each separate subgraph.