A Novel Weighted FP-Stream Algorithm for IoT Data Streams

The Internet of Things (IoT) is a technology that is being widely used in daily life. This technology makes it easier for devices to connect with each other. As a result of the high connectivity between devices, enormous volumes of data are being collected. Such data is called big streaming data which can be used to curate useful information by data mining techniques. One of the most used processing methods is called Frequent Itemset (Pattern) Mining (FIM) which detects recurring and common patterns over data streams. In this paper, a new algorithm based on frequently used FP-Stream algorithm is presented. The proposed algorithm enhances conventional FP-Stream algorithm to make it more adaptive to concept drifts when retaining its applicability to data streams. Conventional FP-Stream algorithms store all detected patterns. By adding weights during the pruning process based on pattern freshness, the proposed algorithm prioritizes newer patterns thereby learns new patterns and forgets older one swiftly. Performance evaluations are performed using data acquired from an IoT testbed established in KAVEM Lab of Gazi University. Evaluation results indicate that the proposed algorithm performs better than conventional FP-Stream significantly.