Friend Recommendation System Based on Mobile Data

The advancement of the mobile technology allows people to use social networking services to connect with each other, share contents, obtain information and make purchases. There are social networks that recommend friends, items, jobs, movies, books… and so on. But, all of them need to constantly build credibility with the active users. These social networking services provide the best recommendation services which focused on the user's main preference' or interest selection. In this paper, we propose a Friend Recommendation system for social networks which will be measure the frequency of user activities and update the dataset according to the frequency and intensity of their activities. We will be use the K-Nearest Neighbor algorithm; we can easily classify the user activity and behaviors with some criteria. We also build a user's daily life journal, from which their life preferences are extracted by using the K-Nearest Neighbors with threshold algorithm. We use a similar method to calculate the similarity of life styles or preference between users, and determine users' impact in terms of interests with a public-private graph. We have implemented the Friend Recommendation system, and evaluated its performance on both small-scale experiments and large-scale simulations. The results show that the recommendations reflect the users 'preferences in choosing friends.