The Design And Implementation Of Web Mining Based On Ant Colony Algorithm

The progress of clustering and classification in web mining was researched in the paper. A classification algorithm based on ant colony optimization was applied in the processing of web page classification in web content mining,moreover, a clustering algorithm based on an improved ant colony algorithm was used in the processing of web user clustering in web usage mining. The result show that, in contrast to traditional method,the clustering and classification algorithm based on ant habit have certain comparative advantages in web mining.First of all in the process of web mining was analysed in this paper, a detailed analysis of web mining clustering and classification of the advantages and disadvantages of existing technology. Discussion of a number of improved ant colony algorithm,after the analysis of the existing web mining algorithm was applied to the technical deficiencies,a novel classification algorithm named Ant_Miner3 based on ant behavior to solve the classification problem during in data mining task was exploited to web page classification of web content mining and the process of non-structure dataset was carried out.Compared with the traditional classification algorithm C5.0, the algorithm Ant_Miner3 can discover more precise and brief rules.A new clustering analysis method, which based on Improved Ant Colony Algorithm, was put forward in this paper, which is named IACA for short. The author realizes the Improved Ant Colony Algorithm (IACA) and design an emulator. With the emulator, the IACA is compared with other clustering analysis algorithms: can avoid the stagnation of the algorithm, avoid the part superior and attain the overall excellent optimization. So it can make the whole capability of IACA to attain optimization.