Research on distributed data stream mining in Internet of things

Abstract

For mining useful data from mass data generated by Internet of things, analyses shortages of the traditional Apriori algorithm which has a lower mining efficiency and occupies the larger memory space. So, MapReduce model of cloud computing is introduced. In the mechanism of MapReduce, combine the architecture characteristics and key technology of Internet of Things, conduct distributedmining on data and information in environment of the Internet of Things, and the calculation model of distributed data stream mining is drawn. Performance analysis proves that the new data stream mining model overcomes difficulty of traditional data mining, and each link reflects the idea of distributed structure and improves mining efficiency obviously.