Social Network Influencers’ Data Augmenting Recommender Systems
Ashrf Althbiti, Xiaogang Ma · 2020
The ever-increasing use of social network sites and the availability of Internet services have introduced opportunities for users to communicate, interact, and connect with one another. Social network sites can identify the similarity between users in order to recommend new friends to a user. Therefore, users might not only communicate with their actual offline friends, but also they are motivated to communicate with strangers and friends of their actual friends. This paper introduces a model that integrates and incorporates social network influencers’ data for augmenting recommender systems. This model is developed based on three main techniques. The first one is the contentbased technique used to recommend items to an active user based on his/her previously collected data or interests. The second technique is a Bayesian classifier used to learn an active user’s profile and to determine social network influencers who can contribute to augment the quality of recommendations. The third technique is used to identify social media influencers. This model can be generalized to other domains that collect a side information from external sources.