The Data Mining Of Iot Based On Cloud Computing

With the continuous development of society,the information industry has gradually become an important pillar of the national economy.As an important part of the new generation of information technology,the Internet of things promotes the development of human civilization and makes the world more intelligent.The data mining of IOT is an important part of IOT technologies and a strong complement to the IOT industry when more and more applications appear in the future.This paper analyzes the difficulties of data mining in IOT because of the data features of IOT,and cloud computing provides an important idea for data mining of IOT.The paper also concludes that cloud computing offers IOT the most powerful computing and storage capacity.In addition,the concept of IOT cloud is described in chapter two.Hadoop which was put forward by Apache is a platform for building cloud.Hadoop makes us easy and fast to build computer clusters.In this paper,there is a detailed description of HDFS and MapReduce programming model which are the two most important technology of Hadoop.Also we introduce distributed storage and distributed data mining and the parallelization of data mining algorithms.Apriori algorithm has been improved through the experiment to verify the feasibility of the whole idea.There are four modules which includes perception layer,transport layer,data storage layer,data mining service layer in our design.And the difficulties and key points for each module are given in details in this article.In addition,during the analysis of the platform,this paper made a research of data converters and open platform interface,which makes the platform more scalability and easy to develop or test for the third party.At present,the applications ecosystem of IOT faces many challenges,and so does the IOT industry chain.This paper gives good ideas to solve these difficulties of IOT.