Research on the Application of Multidimensional Cluster Analysis in Customer Information
AUTHORS
Lu Liu,Harbin Institute of Petroleum, Harbin, Heilongjiang 150027, China
ABSTRACT
Cluster analysis is an important technique in data mining. For products to follow customer needs, it is necessary to make a very precise judgment on the distribution of customers in the market and customer groups. When discussing market development trends, it is necessary to classify and aggregate customer groups in the market. Since customer attribute characteristics are often as high as dozens, data mining clustering algorithms are required to process massive customer historical data and classify and aggregate customer groups, so that companies can develop targeted customer products for different types of customers. This article considers the multi-dimensional customer information, combined with the original idea of the K-means algorithm, realizes the classification and aggregation of a large number of customer information through multi-dimensional aggregation, and compares the performance on Hadoop by comparing the data expansion rate and expansion rate. the expansion rate of the K-means algorithm and the K-means under Hadoop the acceleration ratio of means algorithm parallel operation can find that large batches of data (at least tens of millions of data) are more efficient and accurate in a multi-node cluster Hadoop platform. At the same time, the K-means multi-dimensional attribute clustering algorithm is more suitable for the analysis of customer information data with numerous attributes.
KEYWORDS
Cluster analysis, K-means, Data mining, Hadoop
REFERENCES
[1] X. Huang and Q. Wu, “Micro-blog commercial word extraction based on improved TF-IDF algorithm,” TENCON 2013-2013 IEEE Region 10 Conference (31194), Xi'an, China: IEEE, pp.1-5, (2013)
[2] E. Nikolova and V. Jecheva V. “Some similarity coefficients and application of data mining techniques to the anomaly-based IDS,” Telecommunication Systems, vol.50, no.2, pp.127-135, (2012)
[3] N. Xinzheng and S. Kun, “Research on fast parallel clustering and partitioning algorithm for large-scale data,” Computer Science, vol.39, no.1, pp.134-137, (2012), DOI: 10.3969/j.issn.1002- 137X.2012.01.030.
[4] K. L. Du andM. N. S. Swamy, “Neural networks and statistical learning,” Science Business Media, pp.727-745, (2014)
[5] L. Zhang and Q. Guo, “Research on user clustering method based on multi-dimensional behavior analysis,” Journal of University of Electronic Science and Technology of China, vol.2, pp.315-320, (2020)
[6] J. Liu and H. Guo, “K-means clustering center optimization solution method in cloud computing,” Bulletin of Science and Technology, vol.31, no.10, pp.100-102, (2015)
[7] X. Jiang, C. Li, W. Xiang, “MapReduce parallelization implementation of K-means clustering algorithm,” Journal of Huazhong University of Science and Technology (Natural Science Edition), vol.39, no.1, pp.120-124, (2011)
[8] S. Xiang, F. Nie, and C. S. Zhang, “Learning a Mahalanobis distance metric for data clustering a classification,” Pattern Recognition, (2008)
[9] L. Zeng, “Research on MapReduce parallel computing platform in cloud computing,” Harbin Institute of Technology, (2013)
[10] B. Yang, D. -Y, Liu, and D. Jin, “Complex network clustering algorithms,” Journal of Software, vol.20, no.1, pp.54-66, (2009)