Research of massive searching logs analysis system based on Hive
Zhao Lon · Jisuanji yingyong yanjiu · 2013
Aiming at the bottlenecks of traditional distributed programming model for scalability and parallel programming,this paper proposed a massive Web searching log analysis mechanism based on Hive. By using HQL language,Hadoop distributed file system( HDFS) and MapReduce programming model,it conducted an analysis on user searching behaivors,relevant information taking into consideration including hot keywords in the query,the relationship between user clicks and URL ranking and query sessions. The result of the analysis has a certain guiding significance to the search engine ranking algorithms and system optimization.