Personal social relations research based on Weibo location check-in data

A method for personal social relations research has been proposed in this paper, which is based on the Sina Weibo micro-blogging location check-in data. A framework of mining technology has been designed to analyze the similarities between users by using the spatial and temporal characteristics of their trajectories, which are represented by the Sina Weibo location check-in data. The “user-active areas” are extracted from the weibo location check-in data in order to establish a space vector for each user, on the basis of which the social relations among users could be mined by the similarity of users´ space vectors. Experiments have been conducted to evaluate the proposed method by using the geo location check-in data for Wuhan City. The results show that the method is able to realize friend validation and friend recommendations reliably. It also provides a reference for building social relationship under the Mobile Internet.