ENHANCED E-COMMERCE RECOMMENDATION EXPERIENCE WITH COLLABORATIVE SENTIMENT ANALYSIS AND RANKED CONTENT-BASED FILTERING

Authors

• Sandi El Zein, Department of Mathematics and Computer Science, Faculty of Science, Beirut Arab University, LebanonFollow
• Farah Hamze, Department of Mathematics and Computer Science, Faculty of Science, Beirut Arab University, LebanonFollow
• Amani Merhi, Department of Mathematics and Computer Science, Faculty of Science, Beirut Arab University, LebanonFollow
• Lynn Noureddine, Departement of Mathematics and Computer Science, Faculty of Science, Beirut Arab University, LebanonFollow
• Lama Affara, Department of Mathematics and Computer Science, Faculty of Science, Beirut Arab University, LebanonFollow

Keywords

Recommender Systems, Aspect-Based Sentiment Analysis, Hybrid Recommendation, E-Commerce, Conversational AI

Disciplines

Artificial Intelligence and Robotics | Computer Sciences | Databases and Information Systems

Abstract

The rise of online shopping has made informed purchasing decisions increasingly difficult, as consumers face an overwhelming number of product choices and struggle to manually evaluate specifications and user reviews. This paper presents an AI-powered, sentiment-aware product recommendation tool that effectively aligns user preferences with real-world customer feedback from online reviews. The proposed system utilized Instruct ABSA deep learning models for feature extraction and DeBERTa-v3 for sentiment analysis to turn reviews into interpretable scores that would nominate optimal products. An interactive Rasa-based chatbot interface, TopPickAI, was developed to give a seamless user experience, educate users on product features, and conversationally collect their preferences and priorities through a multi-criteria ranking algorithm. The system was evaluated on a dataset of laptops and reviews for both its effectiveness in ranking devices and user satisfaction of produced recommendations.  Sentiment classification achieved an average accuracy of 94%, outperforming ChatGPT Pro, which achieved 77%.  Furthermore, an 80% majority of survey participants preferred the system’s recommendations over those of ChatGPT Pro. The results verify that fusing aspect-based sentiment analysis, structured filtering, and user-defined priorities yields more personalized and trustworthy product recommendations than general-purpose large language model baselines

Author ORCID Identifier

Lama  Affara www.orcid.org/0000-0001-6950-5347

Recommended Citation

El Zein, Sandi; Hamze, Farah; Merhi, Amani; Noureddine, Lynn; and Affara, Lama (2026) "ENHANCED E-COMMERCE RECOMMENDATION EXPERIENCE WITH COLLABORATIVE SENTIMENT ANALYSIS AND RANKED CONTENT-BASED FILTERING," BAU Journal - Science and Technology: Vol. 7: Iss. 2, Article 3.
DOI: https://doi.org/10.54729/2959-331X.1173

ISSN

2959-331X