A novel parallel clustering algorithm PXM based on FP-Tree

Author

Yonghong Xie ; Chenjun Ling ; Yanhui Ma ; Guoxia Wang

Author_Institution

Sch. of Comput. & Commun. Eng., Univ. of Sci. & Technol. Beijing, Beijing, China

Volume

2

fYear

2012

Firstpage

475

Lastpage

481

Abstract

To solve the following practical problems in e-commerce filed: massive e-commerce data processing, fast clustering of mixed types of data, data sparseness; we present an in-depth discussion and study on the parallel cluster analysis and propose a novel parallel clustering algorithm based on FP-Tree called PXM. In the new algorithm, X-Means Algorithm is improved in two aspects: (1) the usage of MapReduce framework, by taking the appropriate parallel strategy to implement X-Means algorithm. (2) the usage of FP-Tree structure to solve the problem that a collection of type, term type, Boolean data have no definition about mean, easy to use low complexity algorithm like K-Means clustering.

Keywords

data handling; electronic commerce; parallel algorithms; pattern clustering; trees (mathematics); Boolean data; FP-tree; K-means clustering; MapReduce framework; PXM; X-means algorithm; fast clustering; massive e-commerce data processing; parallel cluster analysis; parallel clustering algorithm; Accuracy; Algorithm design and analysis; Clustering algorithms; Computers; Convergence; Educational institutions; Parallel processing; Clustering; Frequent Pattern Tree; MapReduce; Parallelization; X-Means;

fLanguage

English

Publisher

ieee

Conference_Titel

Instrumentation & Measurement, Sensor Network and Automation (IMSNA), 2012 International Symposium on

Conference_Location

Sanya

Print_ISBN

978-1-4673-2465-6

Type

conf

DOI

10.1109/MSNA.2012.6324625

Filename

6324625