high performance computing on graphics processing units: hgpu.org

Text Clustering is the problem of dividing text documents into groups, such that documents in same group are similar to one another and different from documents in other groups. Because of the general tendency of texts forming hierarchies, text clustering is best performed by using a hierarchical clustering method. An important aspect while clustering large text databases is that of high dimensionality of the representation space. Not only does it take lot of space in storing hierarchy trees but also a lot of time is spent in similarity calculations while clustering these documents. In this paper we propose to parallelize a method which uses a tree based summarization technique to store cluster summaries in a tree stored in the memory at all times of processing. The results show that our method shows good accuracy along with a good speed up in calculating clusters.