Consumption Behavior Analysis of E-commerce Users Based on K-means Algorithm

At present, the pattern of China’s e-commerce retail market has been basically formed, Ali already occupies the first place in the market with more than half of the market share, and Jingdong accounts for less than 50 percent, these two companies together account for more than three-quarters of the overall market share in China. However, our country’s national online shopping market is about to become saturated. Therefore, under the premise of the rapid development of the era of big data, the competition of e-commerce platforms has gradually changed from the competition of the number of users to the competition of refined user management. This paper constructs consumer behavior indicators through feature engineering, and then uses the K-means algorithm to cluster customers into four categories, and name these customers as iron powder customers,general customers, develop customers and zombie customers according to their importance to the enterprise from high to low, and analyze the characteristics of different customer groups at the same time, and give corresponding marketing suggestions according to different customer groups. © 2022, Taiwan Ubiquitous Information CO LTD. All rights reserved.