Efficient mining of association rules from Wireless Sensor Networks

Wireless Sensor Networks (WSNs) produce large scale of data in the form of streams. Recently, data mining techniques have received a great deal of attention in extracting knowledge from WSNs data. Mining association rules on the sensor data provides useful information for different applications. Even though there have been some efforts to address this issue in WSNs, they are not suitable when multiple database scans are the major limitation. In this paper, we propose a new tree-based data structure called Sensor Pattern Tree (SP-tree) to generate association rules from WSNs data with one database scan. The SP-tree is constructed in frequency-descending order, which facilitates an efficient mining using the FP-growth-based [6] mining technique. The experimental results show that SP-tree outperforms related algorithms in generating association rules from WSNs data.