Citation:

Mohammad Al Masalma. "Clustering E-Commerce Consumers through Machine Learning-Based Analysis of Clickstream Data" International Research Journal of Economics and Management Studies, Vol. 3, No. 4, pp. 71-77, 2024.

Abstract:

In recent years, the e-commerce sector has significantly grown, with selling strategies becoming increasingly broad and complicated. Businesses aiming to enhance customer experience, optimize marketing efforts, and boost revenue growth must comprehend user behavior. In this research, we examine how well clustering algorithms work to identify significant trends in clickstream data from e-commerce. In order to conclude useful information that can aid in strategic decision-making in a business, we intend to aggregate and analyze user activities such as page visits, product views, and transactions. K-means is a well-known and mostly used clustering algorithm that is used because of its capacity to group data samples according to incommon behavioral patterns. Preprocessed clickstream data, which includes various parameters, including the quantity of clicks, average price viewed, and most popular product, color, location, and model photography, is subjected to each algorithm. We evaluate these algorithms’ performance in terms of cluster quality, interpretability, and significance for e-commerce analytics through extensive testing and comparative research. This paper’s findings identify various user clusters with different product preferences, browsing behaviors, and levels of involvement.

References:

[1] M. Zavali, E. Lacka, and J. De Smedt, “Shopping Hard or Hardly Shopping: Revealing Consumer Segments Using Clickstream Data,” IEEE Trans Eng Manag, vol. 70, no. 4, pp. 1353–1364, Apr. 2023, doi: 10.1109/TEM.2021.3070069.
[2] J.-Z. Lezu, G. Cassani, and A. Hendrickson, “Analyzing and Clustering Clickstream Data for Marketing Personalization,” 2021.
[3] Q. Su and L. Chen, “A method for discovering clusters of e-commerce interest patterns using clickstream data,” Electron Commer Res Appl, vol. 14, no. 1, pp. 1–13, Jan. 2015, doi: 10.1016/j.elerap.2014.10.002.
[4] J. Shlens, “A Tutorial on Principal Component Analysis,” Apr. 2014, [Online]. Available: http://arxiv.org/abs/1404.1100
[5] E. Umargono, J. E. Suseno, and V. Gunawan, “K-Means Clustering Optimization Using the Elbow Method and Early Centroid Determination Based on Mean and Median Formula,” 2020.
[6] S. Nanjundan, S. Sankaran, C. R. Arjun, and G. P. Anand, “Identifying the number of clusters for K-Means: A hypersphere density based approach.”
[7] A. Jain and A. Kumawat, “ANALYSIS OF CLICKSTREAM DATA,” International Research Journal of Engineering and Technology, 2022, [Online]. Available: www.irjet.net

Keywords:

Clickstream Data, Clustering Algorithms, Data analysis, Data mining, K-means, Time series Data.