This study aims to use advanced machine learning models and big data analytics (BDA) to investigate online marketplace client loyalty.To capture nonlinear correlations and identify significant loyalty determinants, the suggested framework uses algorithms like LSTM, XGBoost, support vector machines, and random forests, rather than typical statistical techniques.The study delves into the elements that influence consumer engagement, the frequency of purchases, and the likelihood of repurchases within the fresh food platform and pet-related e-commerce sectors.A ten-phase frameworkcovering foundation, data collection, preprocessing, AI development, pilot study, validation, and evaluation -is employed to ensure methodological rigour.When compared to more traditional methods of consumer loyalty prediction, LSTM performs better on measures including accuracy, precision, recall, F1 score, and area under the curve (AUC).The findings highlight the role of personalised recommendations, delivery services, and mobile engagement in shaping loyalty, while offering practical strategies for sustainable growth in vertical e-commerce markets.
Li et al. (Thu,) studied this question.