Parallel Text Clustering Based on MapReduce
Cao Zewen, Zhou Yao · 2012
This paper analyzes challenges of ordinary text clustering algorithms and proposes cloud computing can be a feasible solution. The classical Jarvis-Patrick (JP) algorithm was adapted as a study case. It was implemented using MapReduce programming mode and was testified on the cloud computing platform-Hadoop with Sogou corpus provided by Sogou laboratory. The experiment results demonstrate that text clustering algorithm can be paralleled in MapReduce framework and parallel algorithm can handle massive textual data and get a better time performance.