Enhancing E-Commerce Recommendation Systems with Deep Learning-based Sentiment Analysis of User Reviews

Abstract

This study introduces a novel approach to enhancing e-commerce recommendation systems by integrating deep learning-based sentiment analysis of user reviews. We propose a sentiment-aware neural collaborative filtering model that leverages the emotional content of reviews to enrich user and item representations. Our method employs a hierarchical attention network for fine-grained sentiment analysis, capturing nuanced user opinions at both word and sentence levels. The sentiment information is then incorporated into a neural collaborative filtering framework, allowing for more personalized and context-aware recommendations. We evaluate our model on a large-scale e-commerce dataset, demonstrating significant improvements in recommendation accuracy, diversity, and user satisfaction compared to state-of-the-art baselines. Our experiments show that the proposed model achieves a 7.66% improvement in NDCG@10 over the strongest baseline, while also enhancing beyond-accuracy metrics such as diversity and novelty. The integration of sentiment analysis proves particularly effective in capturing evolving user preferences and item perceptions, addressing key challenges in traditional recommendation systems. This research contributes to the field by showcasing the potential of leveraging deep learning-based sentiment analysis to create more nuanced, responsive, and user-centric e-commerce recommendation systems.