Title :
Novel Hybrid Document Clustering Algorithm Based on Ant Colony and Agglomerate
Author :
Wang, Xiaohua ; Shen, Jie ; Tang, Hongjun
Author_Institution :
Inst. of Comput. Applic. Technol., Hangzhou Dianzi Univ., Hangzhou, China
fDate :
Nov. 30 2009-Dec. 1 2009
Abstract :
In this paper, ant colony algorithm was improved from two aspects, then a novel hybrid ant colony and agglomerate document clustering algorithm, hybrid-AC&A, has been proposed based on ant colony model and agglomerate clustering algorithms. Firstly, Compact algorithm was applied while ant dropping its load. Secondly, evaluate function based schedule algorithm was applied while ant obtains load. Finally, agglomerate clustering algorithm was integrated into the iteration procedure of ant colony clustering algorithm. The performance of Hybrid-AC&A is compared with other clustering methods, the experimental results denote that the proposed algorithm not only inherits the intrinsic advantages of ant colony model clustering algorithm, but also improves the aspect of time efficiency. Computational result on real documents collection shows it is much more efficient than other mentioned algorithms.
Keywords :
document handling; optimisation; agglomerate clustering algorithm; ant colony algorithm; function based schedule algorithm; hybrid document clustering algorithm; Algorithm design and analysis; Ant colony optimization; Biological system modeling; Clustering algorithms; Clustering methods; Computer applications; Convergence; Density measurement; Knowledge acquisition; Scattering; Agglomerate; Ant Colony; Document Clustering;
Conference_Titel :
Knowledge Acquisition and Modeling, 2009. KAM '09. Second International Symposium on
Conference_Location :
Wuhan
Print_ISBN :
978-0-7695-3888-4
DOI :
10.1109/KAM.2009.182