Study of Data Stream Clustering Based on Bio-inspired Model

Abstract

Abstr act. Nowadays with the rapid development of wireless sensor networks, and network traffic monitoring, stream data gradually becomes one of the most popular data models. Stream data is different from the traditional static data. Clustering analysis is an important technology of data mining, so that many researchers pay their attention to the clustering of stream data. In this paper, MSFS(Multiple Species Flocking on Stream) algorithm is proposed. By means of the experimental verification analysis, MSFS algorithm, which is based on biologically inspired computational model, exists higher clustering purity on both the real dataset and the simulation datasets. In other words, the cluster result of MSFS algorithm is better. Recently, with advances in communication and data collection techniques, people receive a large number of real time data at very high rates. In data mining area, there are many techniques but they should be tuned and changed to work in data stream mining. The data stream mining is different from the regular static data mining. These distinguishing features bring new challenge to stream data processing. Clustering analysis is an important technology of data mining, so that many researchers pay their attention to the clustering of stream data[1]. In this paper, MSFS algorithm is proposed. It combines MSF model and the DenStream clustering algorithm that is based on density. MSF model is a kind of swarm intelligence model for text clustering, and we take advantage of the feature similarity rule to make MSFS be suitable for data stream clustering. This article is organized as follows. The second section describes the related word with the proposed algorithm: the DenStream algorithm and the MSF(Multiple Species Flocking) model. Section 3 describes our algorithm. In 4th section, the results of the method on synthetic and real life data sets are presented. At last section, we discuss the advantages of the approach and concludes this article.