The Algorithm of Data Preprocessing in Web Log Mining Based on Cloud Computing.

In the structure of distributed cluster server, web log data mining model based on data warehouse has the defects of bottlenecks in the network and computing, transmission errors caused by the large data transmission, the paper makes use of the advantages of cloud computing, distributed processing and virtualization technology, designs a type of Web log analysis platform based on cloud computing Hadoop cluster framework, finally, a new hybrid algorithm of distributed procession in the cloud computing environment is proposed. To further verify the efficiency of the platform, we use the improved data pretreatment algorithm on the platform of processing large number of Web logs, experimental results show that it can improve the efficiency of Web data mining. [ABSTRACT FROM AUTHOR]