Biclustering of web usage data using gravitational search algorithm

Author

Diviya Prabha, V. ; Rathipriya, R.

Author_Institution

Dept. of Comput. Sci., Periyar Univ., Salem, India

fYear

2013

fDate

21-22 Feb. 2013

Firstpage

500

Lastpage

505

Abstract

In this paper, an efficient and new algorithm for biclustering of web usage data is presented, which is based on gravitational search algorithm. In the proposed algorithm, called BIC-GSA, the gravitational search algorithm is used to find a near optimal solution for biclustering problem. The benchmark clickstream dataset from UCI repository is used to evaluate and to study the performance of the presented algorithm. The results show that the proposed algorithm can find high quality biclusters in the tested dataset.

Keywords

Internet; data mining; pattern clustering; search problems; software performance evaluation; BIC-GSA; UCI repository; Web usage data; benchmark clickstream dataset; biclustering problem; gravitational search algorithm; high quality biclusters; near optimal solution; performance evaluation; presented algorithm; tested dataset; Algorithm design and analysis; Data mining; Equations; Gravity; Web pages; Biclustering; Correlation based fitness function; Gravitational Search Algorithm (GSA); Web Usage Data; Web Usage Mining;

fLanguage

English

Publisher

ieee

Conference_Titel

Pattern Recognition, Informatics and Mobile Engineering (PRIME), 2013 International Conference on

Conference_Location

Salem

Print_ISBN

978-1-4673-5843-9

Type

conf

DOI

10.1109/ICPRIME.2013.6496722

Filename

6496722