Efficient Diversified Set Monitoring for Mobile Sensor Stream Environments

Due to recent developments in sensor technologies, mobile sensor device use has become widespread, and many researchers have been attempting to leverage data collected by these devices. We call such data 'mobile sensor data'; and environments where mobile sensor data arrive continuously, 'mobile sensor stream' environments. Mobile sensor data are geo-referenced data with environmental attribute values; and they enable us to determine the geographical distribution of hot spots by retrieving data with comparatively extreme environmental attribute values (such as higher air-pollution index values). Top-k search result diversification in geographical space is valid for applications of this sort. By monitoring a diversified set over mobile sensor streams, we can trace changes in the distribution of hot spots. However, the computation costs for maintaining such diversified sets are high when we have to monitor a large amount of mobile sensor data. Thus, in this paper, we propose an efficient diversified set monitoring method for mobile sensor stream environments. Our proposed method can reduce the amount of examined data by exploiting our proposed regular grid-based data structure, and the diversified set can thereby be maintained much more efficiently. Our experimental results confirm that the proposed method involves much shorter computation time in comparison with the baseline method.