2 January 2024 in IEEE Access
With the growth of the internet and e-commerce, online reviews have become a prevalent and rich source of information for personalized recommendations. Review text typically contains user preferences regarding various aspects of items or services. However, most previous recommendation models that used online reviews focused on general opinions, without considering the various aspects of the review text. Although previous approaches have effectively reflected the overall preferences embedded in reviews, capturing the preferences for various aspects is limited. This paper proposes an attentive aspect-based recommendation model with a deep neural network (AARN), which can capture user preferences regarding various aspects embedded in review text. The proposed model uses an advanced aspect-based sentiment analysis (ABSA) method. Unlike previous approaches that extract overall preferences, ABSA offers the advantage of extracting preferences for various aspects embedded in review texts. An attention mechanism is then applied to measure the unique attention weights for each aspect. Extensive experiments were conducted on real-world online review datasets, showing that the proposed model outperformed baseline models in effectively predicting user ratings for target items. Furthermore, this study demonstrated the influence of the ABSA method on personalized and explainable recommendations.
25 April 2024 in IEEE Access
The expansion of the e-commerce market has led to the challenge of information overload, necessitating the development of recommender systems. The recommender system aids users in decision-making by suggesting items that align with their preferences. However, conventional recommendation models rely solely on quantitative user behavior data, such as user ratings, and lead to limitations in recommendation performance due to the sparsity problem. To address these issues, recent research has leveraged convolutional neural networks (CNNs) to extract and incorporate semantic information from user reviews. However, several prior studies have a disadvantage in that they fail to account for the intricate interactions between users and items directly. In this study, we introduce a novel approach, the Review-based recommender system using Outer Product on CNN (ROP-CNN) model, which adeptly captures and incorporates semantic features from reviews to address the complex interactions between users and items using CNN. The experimental results, using real user-review datasets, demonstrate that the ROP-CNN model outperforms existing baseline models for prediction accuracy. And this study presents a novel theoretical and methodological perspective in recommendation research, suggesting a method that integrates user preference information from reviews into recommender systems by leveraging rich user-item interaction information.