Accurate prediction of purchasing behaviour of cross border e-commerce consumers under social media marketing

Traditional cross-border e-commerce consumer purchase behaviour prediction methods have problems with low reliability and high prediction error rate. This article designs a research on accurate prediction of purchasing behaviour of cross border e-commerce consumers under social media marketing. Firstly, determine and extract the purchasing behaviour characteristics of cross-border e-commerce consumers under social media marketing. Then, determine the initial centroid set of feature data, determine similar data by calculating the Euclidean distance between feature data, and remove similar data. Finally, by calculating the information entropy of the feature data, determine the weight value of the feature data, and use the integration algorithm and loss function to achieve accurate prediction of purchase behaviour. The test results show that the proposed method improves the reliability of prediction and reduces the prediction error rate. © 2024 Inderscience Enterprises Ltd.