A Group Mining Method based on Personal Interaction in Mobile Environment

Group mining is to identifying meaningful groups from users. Most of the existing group mining methods create groups based on physical information like spatio-temporal information. Other approaches are based on social relation such as user interaction or common preferences. However, methods using only physical information create groups which are simply comprised of adjacent people to a certain user. In addition, once methods using only social information create a group, the group is fixed by a certain social factor. In this paper, we propose a grouping method conflating physical information and social information collected from user’s mobile device to solve this problem. The proposed method is appropriate to create a group for moving users. This method reflects mobility of users and sensitivity of social situation at the same time.