With the rapid expansion of e-commerce platforms, the demand for highly accurate and personalized recommendation systems has become increasingly prominent. Traditional recommendation algorithms often struggle to capture the complex and dynamic nature of user preferences, especially when dealing with heterogeneous data sources. In this study, we propose a novel recommendation framework that synergistically integrates user clustering, BERT-based sentiment analysis, contextual encoding, and deep learning techniques. Leveraging a real-world dataset from Kaggle, the proposed model comprehensively incorporates user behavior records, review texts, and contextual information to construct rich user and item representations. Dimensionality reduction and clustering methods are employed to identify latent user groups, while BERT is utilized to extract deep semantic features from user-generated reviews. The fused feature vectors are then fed into a multi-layer perceptron to generate personalized recommendations. Extensive experiments demonstrate that the proposed K-means+BERT+MLP model consistently outperforms a wide range of traditional and hybrid baselines across multiple evaluation metrics, including accuracy, precision, recall, F1-score, and AUC. The results validate the effectiveness and robustness of the proposed approach, highlighting the potential of multi-source feature fusion and advanced modeling techniques in next-generation recommender systems.
Li et al. (Wed,) studied this question.