Attention long short-term memory-based website analysis on the impact of cross-border electronic commerce on online consumer behavior

Good website design can guide consumers more convenient to find the desired goods, more easily familiar with the website function, more convenient for a series of operations. But the existing website design relies on the traditional method, cannot recommend the real suitable products for consumers. This paper proposes an Attention Long Short-Term Memory (ALSTM) model designed to enhance website service quality in cross-border e-commerce by effectively analyzing and predicting consumer behavior. The ALSTM model leverages LSTM layers to capture the sequential patterns of user click behaviors, while an attention mechanism is applied to prioritize relevant product attributes based on consumer interactions. By dynamically adjusting to consumer preferences, the model provides personalized product recommendations, ultimately improving user engagement and purchase potential. Experimental results demonstrate that the ALSTM model outperforms baseline algorithms, achieving substantial improvements in mean absolute error (MAE) and root mean square error (RMSE) on the Ciao dataset, underscoring its efficacy in delivering accurate and adaptive recommendations.