Research of Elevator Security Data Mining Based on Hadoop

Abstract

Hadoop data platform is studied to mine the massive elevator data from the elevator remote supervision system. The data mining and analysis platform is designed by Hadoop minimum cluster theory, which is mainly four modules composed of data transmission, data preprocessing, data mining and scheduling. Then both K-Means which is a cluster analysis algorithm and Apriori which is an association rule algorithm are improved, and parallelized with the computation framework of Hadoop's MapReduce. Finally, all modules are integrated through the scheduling module, and the practicability of the data mining and analysis platform is verified and tested, and the data mining of security status is realized for the elevator.