Research on quantitative measurement algorithm for e-commerce customer loyalty based on deep learning algorithm

Traditional algorithms cannot fully explore the potential patterns behind big data, lack personalized customer analysis, and cannot provide personalized services and suggestions for different types of customers. This article employs the Bi LSTM (Bidirectional Long Short-Term Memory) model to accurately capture the complex features and patterns of customer behavior, thereby improving the measurement accuracy of customer loyalty. Collect data on customer behavior, browsing history, and search behavior, and preprocess the collected data. Organize customer behavior data into a time series dataset in chronological order, and divide it into weekly windows to extract feature information from the data. Construct a bidirectional LSTM model while considering the forward and backward information of the sequence data, in order to more comprehensively capture the contextual relationships in the sequence data and quantify customer loyalty. The experimental results show that the average accuracy of Bi LSTM in predicting average customer loyalty is 97.1%. And it can effectively improve the prediction effect of repeat purchase rate. The application of Bi LSTM can accurately quantify customer loyalty in e-commerce, provide reference for enterprise decision-making, formulate corresponding marketing strategies and customer management plans, and improve customer loyalty and competitive advantages.