在這資訊科技快速發展的年代，降低了企業對於資料蒐集的門檻，企業與
顧客的連結更加緊密，因此瞭解並掌握顧客儼然已是現今行銷人員最重要的課
題，以及智慧零售透過數據分析顧客全通路的購物、瀏覽行為，使企業針對不
同屬性的會員，提供精準的行銷資訊，因此顧客分群成為重要方法，目的是將
顧客依照不同的特性劃分，大部分的行銷人員以 RFM 模型和集群分析來完成
顧客分群，然而此方法依然存在許多可改進的地方，且鮮少將交易資料以外的
數據也納入分析中，因此本研究欲應用多種機器學習方法以及多個面向的變數
進行顧客的分群分析。
本研究透過台灣零售業的資料進行顧客分群分析，採用極限梯度提升樹
(XGBoost)找出關鍵的特徵值，接著透過多種集群分析演算法的比較找出最合適
的分群方法與結果，包含 K-means、Hierarchical Clustering、Birch 和 SOM，後
續針對不同群體的各個變數進行討論，接著進行關聯規則分析，除了購物籃分
析之外本研究也探討了網路瀏覽和促銷之間的關聯性。本研究結果在 RFM 模
型的三個變數之外，另外找出兩個重要變數彌補了原有模型的缺陷，且 Birch
演算法在該資料的集群分析具有良好的表現，最後將顧客分為五群，針對分群
結果進一步的探討分析，深入了解各集群的變數差異性以及顧客行為的不同，
並透過關聯規則分析找出多個關聯組合，為企業提供決策支援並使顧客價值最
大化。

With the rapid development of information technology, companies can build
connections with customers more easily by collecting various data. Therefore,
understanding customers has become the most important issue for today's marketers.
By using data to analyze customers' shopping and browsing behaviors in all channels,
retail companies can provide accurate marketing information for different customers.
Therefore, how to divide customers according to different features and complete
customer segmentation has become one of the most important research topics in retail.
While most marketers use RFM model and cluster analysis to complete customer
segmentation, the drawback of this method is that it only considers recent purchase
and only have transaction data in the analysis. For this reason, this study aims to apply
different machine learning methods with a variety of variables for customer
segmentation analysis.
In this thesis, customer segmentation analysis was conducted based on the data
of Taiwan's retail industry. I use extreme gradient boosting tree (XGBoost) to find the
important features and the most suitable clustering method and results through the
comparison of various cluster analysis algorithms, including K- means, Hierarchical
Clustering, Birch, and SOM. Finally, I discuss the impact of various variables for
different groups by association rule analysis. The results of the research show that
besides the three features of the RFM model, there are two important variables such
as Period and NES that could improve the defects of the original model, and Birch
algorithm has a good performance in the cluster analysis of the data. The results of
clustering analysis indicate that customers can be divided into five segments with
different features and purchase behaviors. Together with the results of clustering
analysis and association rules, retailers can design suitable marketing campaign and
improve their customer relationships.

Agrawal, R., Imieliński, T., & Swami, A. (1993). Mining association rules between
sets of items in large databases. Paper presented at the Proceedings of the
1993 ACM SIGMOD international conference on Management of data.
Agrawal, R., & Srikant, R. (1994). Fast algorithms for mining association rules.
Paper presented at the Proc. 20th int. conf. very large data bases, VLDB.
Chen, D., Sain, S. L., Guo, K. J. J. o. D. M., & Management, C. S. (2012). Data
mining for the online retail industry: A case study of RFM model-based
customer segmentation using data mining. Journal of Database Marketing &
Customer Strategy Management, 19(3), 197-208.
Chen, M., Liu, Q., Chen, S., Liu, Y., Zhang, C.-H., & Liu, R. J. I. A. (2019).
XGBoost-based algorithm interpretation and application on post-fault transient
stability status prediction of power system. IEEE Access, 7, 13149-13158.
Cheng, C.-H., & Chen, Y.-S. J. E. s. w. a. (2009). Classifying the segmentation of
customer value via RFM model and RS theory. Expert systems with
applications, 36(3), 4176-4184.
Dogan, O., Ayçin, E., Bulut, Z. A. J. I. J. o. C. E., & Sciences, A. (2018). Customer
segmentation by using RFM model and clustering methods: a case study in
retail industry. International Journal of Contemporary Economics and
Administrative Sciences, 8(1), 1-19.
Hughes, A. M. (1994). Strategic database marketing : the masterplan for starting and
managing a profitable, customer-based marketing program.
Kaur, M., & Kang, S. J. P. c. s. (2016). Market Basket Analysis: Identify the changing
trends of market data using association rule mining. Procedia computer
science, 85, 78-85.
Kohonen, T., & Honkela, T. J. S. (2007). Kohonen network. 2(1), 1568.
Kotler, P. (1997). Marketing management : analysis, planning, implementation and
control. Upper Saddle River (New Jersey); London: Prentice Hall
International : Prentice-Hall.
Kotler, P., & Armstrong, G. (2010). Principles of marketing: Pearson education.
Kumar, T. S. J. J. o. A. I. (2020). Data mining based marketing decision support
system using hybrid machine learning algorithm. Journal of Artificial
Intelligence, 2(03), 185-193.
Kuo, R. J., Ho, L., Hu, C. J. C., & Engineering, I. (2002). Cluster analysis in
industrial market segmentation through artificial neural network. Computers &
Industrial Engineering, 42(2-4), 391-399.
Lisheng, L., & Ming, L. (2020). Research on the Training Mode of Applicationoriented Marketing Talents under the Smart New Retail. Paper presented at the
2020 International Conference on Big Data and Informatization Education
(ICBDIE).
MacQueen, J. B. (1965). On the Asymptotic Behavior of k-means.
Marcus, C. (1998). A practical yet meaningful approach to customer segmentation.
Journal of Consumer Marketing, 15(5), 494-504.
doi:10.1108/07363769810235974
Peker, S., Kocyigit, A., Eren, P. E. J. M. I., & Planning. (2017). LRFMP model for
customer segmentation in the grocery retail industry: a case study.
Punj, G., & Stewart, D. W. J. J. o. m. r. (1983). Cluster analysis in marketing research:
Review and suggestions for application. Journal of marketing research, 20(2),
134-148.
Rong, J., Vu, H. Q., Law, R., & Li, G. J. T. M. (2012). A behavioral analysis of web
sharers and browsers in Hong Kong using targeted association rule mining.
Tourism Management, 33(4), 731-740.
Rousseeuw, P. J. J. J. o. c., & mathematics, a. (1987). Silhouettes: a graphical aid to
the interpretation and validation of cluster analysis. Journal of computational
and applied mathematics, 20, 53-65.
Shaw, M. J., Subramaniam, C., Tan, G. W., & Welge, M. E. J. D. s. s. (2001).
Knowledge management and data mining for marketing. 31(1), 127-137.
Smith, W. R. J. J. o. m. (1956). Product differentiation and market segmentation as
alternative marketing strategies. Journal of marketing, 21(1), 3-8.
Sokol, O., & Holý, V. J. I. J. o. M. R. (2021). The role of shopping mission in retail
customer segmentation. International Journal of Market Research, 63(4), 454-
470.
Taylor, R. J. J. o. d. m. s. (1990). Interpretation of the correlation coefficient: a basic
review. Journal of diagnostic medical sonography, 6(1), 35-39.
Tedlow, R. S. (2014). The fourth phase of marketing: marketing history and the
business world today. In The Rise and Fall of Mass Marketing (RLE
Marketing) (pp. 24-51): Routledge.
Zhang, T., Ramakrishnan, R., & Livny, M. J. A. s. r. (1996). BIRCH: an efficient data
clustering method for very large databases. ACM sigmod record, 25(2), 103-
114.
Zheng, H., Yuan, J., & Chen, L. J. E. (2017). Short-term load forecasting using EMDLSTM neural networks with a Xgboost algorithm for feature importance
evaluation. Energies, 10(8), 1168.
Zheng, Z., Kohavi, R., & Mason, L. (2001). Real world performance of association
rule algorithms. Paper presented at the Proceedings of the seventh ACM
SIGKDD international conference on Knowledge discovery and data mining.