Affiliations: [a] College of Science, Huazhong Agricultural University, Wuhan, Hubei, China
| [b] School of Economics, Yonsei University, Seoul, South Korea
Correspondence:
[*]
Corresponding author. Xinglong Ren, College of Science, Huazhong Agricultural University, Wuhan, Hubei 430070, China. E-mail: r-xinglong@mail.hzau.edu.cn.
Abstract: Currently, the fuzzy clustering algorithm of customer group behavior data had the poor effect of data clustering. Therefore, a fuzzy clustering algorithm of Internet customer group behavior data based on fuzzy C means clustering was proposed. By constructing the feature vector of behavior data, this algorithm realized the feature extraction of behavior data, and then it used the nearest neighbor chain to extract data features for the reduction and sample equilibrium. The classification of Internet customer group behavior data was achieved. According to the fuzzy C means clustering algorithm, the interval estimation of classification results of behavior data was carried out. Meanwhile, the membership values of each data sample were updated. Finally, the classification interval was adjusted. Thus, the fuzzy clustering of Internet user group behavior data was completed. Experiment results show that the proposed algorithm has high accuracy in data classification, short execution time in clustering, less memory footprint and low computational complexity, which improves clustering effect.
