Authors: Chong-Huan Xu
Addresses: Business Administration College, Contemporary Business and Trade Research Center, Contemporary Business and Collaborative Innovation Research Center, Zhejiang Gongshang University, Hangzhou City, China
Abstract: This paper presents a novel approach to effectively clustering a large amount of data stream produced by some applications such as large-scale surveillance, network packet inspection and stock market. Owing to the massiveness and forgotten characteristics of the data stream, the proposed approach uses a damped window model to partition them. Then it adopts modified K-means based on the Artificial Bee Colony (ABC) algorithm to cluster this data stream fragment and dynamically update the clustering result. Detailed simulation analysis demonstrates that this algorithm is of high efficiency of space and time and is more stable.
Keywords: damped window model; artificial bee colony; ABC algorithm; modified K-means clustering; data stream clustering; simulation; large-scale surveillance; network packet inspection; stock markets.
International Journal of Wireless and Mobile Computing, 2015 Vol.8 No.1, pp.59 - 65
Received: 17 Jul 2014
Accepted: 20 Aug 2014
Published online: 02 Jan 2015 *