Entre Nota e Opinião: Análise de Sentimento em Reviews para Recomendação Colaborativa

• Naan Vasconcelos UFSJ
• Yan Andrade UFMG
• Elisa Tuler UFSJ
• Frederico A. Durão UFBA
• Adriano C. M. Pereira UFMG
• Leonardo Rocha UFSJ

DOI: https://doi.org/10.5753/webmedia.2026.30191

Resumo

Collaborative recommender systems often treat explicit ratings as reliable signals of user preference. However, a numerical score may compress complex experiences and fail to reflect the opinion expressed in a textual review. This paper investigates the divergence between ratings and reviews and analyzes whether sentimentinferred scores can be used to semantically recalibrate the feedback matrix used by collaborative recommendation models.We apply three sentiment analysis strategies, a lexical approach (VADER) and two approaches based on pre-trained language models (BERT and BART) to infer alternative scores from reviews and compare them with the original ratings. We then replace the original ratings with sentiment-inferred scores and evaluate four collaborative algorithms: UserKNN, ItemKNN, SVD, and EASE. Experiments are conducted on four datasets from two domains, e-commerce and Points of Interest (POI), using two Amazon collections and two Yelp subsets. The results showthat rating–review divergences occur in all datasets and are more pronounced in collections with longer reviews. Replacing original ratings with sentiment-inferred scores improves recommendation effectiveness across several scenarios, especially for global collaborative models such as SVD and EASE, whereas neighborhood-based methods exhibit less stable behavior. However, these gains often come at the expense of diversity and novelty, revealing a trade-off between semantic feedback alignment and recommendation variety. Overall, this study shows that textual reviews should not be treated only as auxiliary features for predicting ratings, but also as independent signals that can question, complement, or recalibrate explicit feedback in collaborative recommendation.

Palavras-chave: recommender systems, sentiment analysis, collaborative filtering, reviews, rating-review divergence, beyond-accuracy evaluation

Referências

Gediminas Adomavicius and Alexander Tuzhilin. 2005. Toward the next generation of recommender systems: A survey of the state-of-the-art and possible extensions. IEEE transactions on knowledge and data engineering 17, 6 (2005), 734–749.

Charu C. Aggarwal. 2016. Recommender Systems: The Textbook (1 ed.). Springer, Cham. DOI: 10.1007/978-3-319-29659-3

Sumaia Mohammed Al-Ghuribi and Shahrul Azman Mohd Noah. 2019. Multi-Criteria Review-Based Recommender System–The State of the Art. IEEE Access 7 (2019), 169446–169468. DOI: 10.1109/ACCESS.2019.2954861

Amal Almansour, Reem Alotaibi, and Hajar Alharbi. 2022. Text-Rating Review Discrepancy (TRRD): An Integrative Reviewand Implications for Research. Future Business Journal 8, 1 (2022), 3. DOI: 10.1186/s43093-022-00114-y

Xavier Amatriain and Justin Basilico. 2015. Recommender Systems in Industry: A Netflix Case Study. 385–419. DOI: 10.1007/978-1-4899-7637-6_11

Konstantin Bauman, Bing Liu, and Alexander Tuzhilin. 2017. Aspect Based Recommendations: Recommending Items with the Most Valuable Aspects Based on User Reviews. In Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (Halifax, NS, Canada) (KDD ’17). Association for Computing Machinery, New York, NY, USA, 717–725. DOI: 10.1145/3097983.3098170

Guilherme Bittencourt, Guilherme Fonseca, Yan Andrade, Nícollas Silva, and Leonardo Rocha. 2023. A Survey on Review - Aware Recommendation Systems. In Proceedings of the 29th Brazilian Symposium on Multimedia and the Web (Ribeirão Preto, Brazil) (WebMedia ’23). 198–207. DOI: 10.1145/3617023.3617050

Guilherme Bittencourt, Naan Vasconcelos, Yan Andrade, Nicollas Silva, Washington Cunha, Diego Roberto Colombo Dias, Marcos André Gonçalves, and Leonardo Rocha. 2025. Aware Recommender Systems (RARSs): Recent Advances, Experimental Comparative Analysis, Discussions, and New Directions. Comput. Surveys 58, 1 (2025), 1–49.

Guilherme Bittencourt, Naan Vasconcelos, and Leonardo Rocha. 2025. iRev: Um framework de avaliação de sistemas de recomendação baseados comentários textuais. Revista Eletrônica de Iniciação Científica em Computação 23 (2025), 130–135.

Erik Cambria, Soujanya Poria, Alexander Gelbukh, and Mike Thelwall. 2017. Sentiment analysis is a big suitcase. IEEE intelligent systems 32, 6 (2017), 74–80.

Oscar Celma and Paul Lamere. 2011. Music recommendation and discovery revisited. In Proceedings of the Fifth ACM Conference on Recommender Systems (Chicago, Illinois, USA) (RecSys ’11). Association for Computing Machinery, New York, NY, USA, 7–8. DOI: 10.1145/2043932.2043936

Chien Chin Chen, Shun-Yuan Shih, and Meng Lee. 2016. Who should you follow? Combining learning to rank with social influence for informative friend recommendation. Decision Support Systems 90 (2016), 33–45. DOI: 10.1016/j.dss.2016.06.017

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 conference of the North American chapter of the association for computational linguistics: human language technologies, volume 1 (long and short papers). 4171–4186.

Nan Hu, Paul A Pavlou, and Jennifer Zhang. 2006. Can online reviews reveal a product’s true quality? Empirical findings and analytical modeling of online word-of-mouth communication. In Proceedings of the 7th ACM conference on Electronic commerce. 324–330.

Clayton Hutto and Eric Gilbert. 2014. Vader: A parsimonious rule-based model for sentiment analysis of social media text. In Proceedings of the international AAAI conference on web and social media, Vol. 8. 216–225.

Yehuda Koren, Robert Bell, and Chris Volinsky. 2009. Matrix factorization techniques for recommender systems. Computer 42, 8 (2009), 30–37.

Yehuda Koren, Steffen Rendle, and Robert Bell. 2021. Advances in collaborative filtering. Recommender systems handbook (2021), 91–142.

Balraj Kumar and Neeraj Sharma. 2016. Approaches, Issues and Challenges in Recommender Systems: A Systematic Review. Indian Journal of Science and Technology 9 (12 2016). DOI: 10.17485/ijst/2015/v8i1/94892

Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Veselin Stoyanov, and Luke Zettlemoyer. 2020. BART: Denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehension. In Proceedings of the 58th annual meeting of the association for computational linguistics. 7871–7880.

Greg Linden, Brent Smith, and Jeremy York. 2003. Amazon. com recommendations: Item-to-item collaborative filtering. IEEE Internet computing 7, 1 (2003), 76–80.

Washington Luiz, Felipe Viegas, Rafael Alencar, Fernando Mourão, Thiago Salles, Dárlinton Carvalho, Marcos Andre Gonçalves, and Leonardo Rocha. 2018. A Feature-Oriented Sentiment Rating for Mobile App Reviews. In Proceedings of the 2018 World Wide Web Conference (Lyon, France) (WWW ’18). InternationalWorld Wide Web Conferences Steering Committee, Republic and Canton of Geneva, CHE, 1909–1918. DOI: 10.1145/3178876.3186168

Emanuelle Marreira, Tiago Melo, Miguel Oliveira, and Carlos Maurício. 2025. Detectando Incoerências Avaliativas em E-commerce com LLMs - Um Estudo de Caso na Amazon Brasil. In Proceedings of the 31st Brazilian Symposium on Multimedia and theWeb (Rio de Janeiro/RJ). SBC, Porto Alegre, RS, Brasil, 535–539. DOI: 10.5753/webmedia.2025.15952

Jaewan Moon, Hye-young Kim, and Jongwuk Lee. 2023. It’s Enough: Relaxing Diagonal Constraints in Linear Autoencoders for Recommendation. In Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval (Taipei, Taiwan) (SIGIR ’23). Association for Computing Machinery, New York, NY, USA, 1639–1648. DOI: 10.1145/3539618.3591704

Bo Pang and Lillian Lee. 2008. Opinion mining and sentiment analysis. Now Publishers Inc.

Paul Resnick, Neophytos Iacovou, Mitesh Suchak, Peter Bergstrom, and John Riedl. 1994. Grouplens: An open architecture for collaborative filtering of netnews. In Proceedings of the 1994 ACM conference on Computer supported cooperative work. 175–186.

Badrul Sarwar, George Karypis, Joseph Konstan, and John Riedl. 2001. Itembased collaborative filtering recommendation algorithms. In Proceedings of the 10th International Conference on World Wide Web (Hong Kong, Hong Kong) (WWW’01). Association for Computing Machinery, New York, NY, USA, 285–295. DOI: 10.1145/371920.372071

Badrul Sarwar, George Karypis, Joseph Konstan, and John Riedl. 2001. Item-based collaborative filtering recommendation algorithms. In Proceedings of the 10th international conference on World Wide Web. 285–295.

J. Ben Schafer, Dan Frankowski, Jon Herlocker, and Shilad Sen. 2007. Collaborative Filtering Recommender Systems. Springer Berlin Heidelberg, Berlin, Heidelberg, 291–324. DOI: 10.1007/978-3-540-72079-9_9

Alex Shtoff, Michael Viderman, Naama Haramaty-Krasne, Oren Somekh, Ariel Raviv, and Tularam Ban. 2024. Low Rank Field-Weighted Factorization Machines for LowLatency Item Recommendation. In Proceedings of the 18th ACMConference on Recommender Systems (Bari, Italy) (RecSys ’24). Association for Computing Machinery, New York, NY, USA, 238–246. DOI: 10.1145/3640457.3688097

Rodrigo Souza and Marcelo Manzato. 2024. Uma Abordagem em Etapa de Processamento para Redução do Viés de Popularidade. In Proceedings of the 30th Brazilian Symposium on Multimedia and the Web (Juiz de Fora/MG). SBC, Porto Alegre, RS, Brasil, 310–317. DOI: 10.5753/webmedia.2024.241542

Mehdi Srifi, Ahmed Oussous, Ayoub Ait Lahcen, and Salma Mouline. 2020. Recommender Systems Based on Collaborative Filtering Using Review Texts—A Survey. Information 11, 6 (2020). DOI: 10.3390/info11060317

Harald Steck. 2019. Embarrassingly shallow autoencoders for sparse data. In The World Wide Web Conference. 3251–3257.

Fei Sun, Jun Liu, Jian Wu, Changhua Pei, Xiao Lin, Wenwu Ou, and Peng Jiang. 2019. BERT4Rec: Sequential recommendation with bidirectional encoder representations from transformer. In Proceedings of the 28th ACM international conference on information and knowledge management. 1441–1450.

Naan Vasconcelos, Davi Reis, Thiago Silva, Nícollas Silva, Washington Cunha, Elisa Tuler, and Leonardo Rocha. 2024. Integrando Avaliações Textuais de Usuários em Recomendação baseada em Aprendizado por Reforço. In Proceedings of the 30th Brazilian Symposium on Multimedia and the Web (Juiz de Fora/MG). SBC, Porto Alegre, RS, Brasil, 390–394. DOI: 10.5753/webmedia.2024.241405

Joao Vinagre, Alípio Mário Jorge, Conceição Rocha, and Joao Gama. 2019. Statistically robust evaluation of stream-based recommender systems. IEEE Transactions on Knowledge and Data Engineering 33, 7 (2019), 2971–2982.

Wenpeng Yin, Jamaal Hay, and Dan Roth. 2019. Benchmarking zero-shot text classification: Datasets, evaluation and entailment approach. In Proceedings of the 2019 conference on empirical methods in natural language processing and the 9th international joint conference on natural language processing (EMNLP-IJCNLP). 3914–3923.

Shuai Zhang, Lina Yao, Aixin Sun, and Yi Tay. 2019. Deep Learning Based Recommender System: A Survey and New Perspectives. ACM Comput. Surv. 52, 1, Article 5 (Feb. 2019), 38 pages. DOI: 10.1145/3285029

Yongfeng Zhang, Guokun Lai, Min Zhang, Yi Zhang, Yiqun Liu, and Shaoping Ma. 2014. Explicit factor models for explainable recommendation based on phrase-level sentiment analysis. In Proceedings of the 37th International ACM SIGIR Conference on Research & Development in Information Retrieval (Gold Coast, Queensland, Australia) (SIGIR ’14). Association for Computing Machinery, New York, NY, USA, 83–92. DOI: 10.1145/2600428.2609579