Identifying the Space Buddies to Track Lost Items

Locating missing or lost objects has always been a challenging task. RFID technology and participatory sensing based approaches have offered solutions but often their adoption was limited due to the high hardware costs or low active participation problem. With the introduction of iBeacon technology and smartphones having BLE capability, tracking such objects has become easier and cost-effective. Objects of care are labeled by attaching to them affordable iBeacon tags, and smartphones in the proximity of these tags sense their presence opportunistically through the applications running in the background. In this paper, we study the tracking of lost objects through the collaboration among users. We analyze the visit patterns of users at the same locations and develop a metric that quantifies for each user the potential benefit of others in terms of their capability of finding that user's lost objects. Depending on the predicted benefits, each user's preference list of other users is formed and then utilized to identify the space buddies who can best track her lost items. The identification is based on the adaption of the solution to the roommate matching problem. We apply the proposed system to two different location based social network datasets and show its effectiveness in different settings.