Detecting Sophisticated Fake Reviews on E-Commerce Platforms Using Adversarial Transformer Networks
Received: 14 July 2025 | Revised: 12 September 2025 and 6 October 2025 and | Accepted: 9 October 2025 | Online: 8 December 2025
Corresponding author: Sabar Aritonang Rajagukguk
Abstract
The proliferation of Artificial Intelligence (AI)-generated fake reviews poses an unprecedented threat to the integrity of e-commerce platforms, particularly in developing markets where regulatory frameworks remain nascent. This study proposes an adversarial transformer network framework specifically designed to detect sophisticated fake reviews on Indonesian e-commerce platforms. We developed a novel adversarial training architecture that pairs a Bidirectional Encoder Representations from Transformers (BERT)-based classifier model with a generator capable of producing human-like fake reviews, creating an iterative optimization process that enhances detection robustness. The scientific novelty of this work is threefold: (i) architectural innovation, through the integration of IndoBERT as a discriminator with a fine-tuned Generative Pre-trained Transformer (GPT)-based generator in a competitive adversarial loop; (ii) linguistic innovation, by embedding Indonesian-specific preprocessing (slang handling, code-mixed normalization, emoticon filtering) to address multilingual and culturally diverse contexts; and (iii) training innovation, by introducing gradient penalty mechanisms and iterative adversarial updates that enhance robustness against Large Language Model (LLM)-generated reviews. Together, these contributions distinguish our framework from prior adversarial Natural Language Processing (NLP) approaches that primarily focused on English-language data and lacked local linguistic customization. To the best of our knowledge, this represents the first adversarial transformer framework tailored for Indonesian e-commerce fake review detection. Using a comprehensive dataset of 50,000 authentic reviews collected from major Indonesian e-commerce platforms (Tokopedia and Shopee) and 25,000 AI-generated fake reviews, our methodology achieved significant improvements over traditional detection methods. The adversarial framework demonstrated superior performance with an accuracy of 94.3%, precision of 93.8%, recall of 94.7%, and F1-score of 94.2%, outperforming baseline BERT models by 8.7% in accuracy. Our approach addresses the critical challenge of detecting increasingly sophisticated AI-generated fake reviews while providing insights into the unique linguistic patterns of Indonesian online commerce discourse. The findings contribute to both the theoretical understanding of adversarial learning in NLP and practical applications for maintaining trust in digital marketplaces.
Keywords:
adversarial networks, fake review detection, Bidirectional Encoder Representations from Transformers (BERT), e-commerce, Indonesian market, transformer models, Natural Language Processing (NLP), digital trustDownloads
References
S. Sankhla and A. Katiyar, "The Influence of Online Reviews on Consumer Learning and Purchase Decisions," International Research Journal on Advanced Engineering and Management, vol. 2, no. 11, pp. 3427–3430, Nov. 2024. DOI: https://doi.org/10.47392/IRJAEM.2024.0504
F. L. Witi and A. Mude, "Implementasi Web E-Commerce Berbasis Content Management System Wordpress pada DND Komputer," Jupiter, vol. 16, no. 2, pp. 701–712, Sep. 2024.
J. Thevakumar and L. Thevakumar, "RATHAN@DravidianLangTech 2025: Annaparavai - Separate the Authentic Human Reviews from AI-generated one," in Proceedings of the Fifth Workshop on Speech, Vision, and Language Technologies for Dravidian Languages, Albuquerque, NM, USA, 2025, pp. 449–453. DOI: https://doi.org/10.18653/v1/2025.dravidianlangtech-1.66
P. Hajek and J.-M. Sahut, "Mining behavioural and sentiment-dependent linguistic patterns from restaurant reviews for fake review detection," Technological Forecasting and Social Change, vol. 177, Apr. 2022, Art. no. 121532. DOI: https://doi.org/10.1016/j.techfore.2022.121532
M. A. Wani, M. ElAffendi, and K. A. Shakil, "AI-Generated Spam Review Detection Framework with Deep Learning Algorithms and Natural Language Processing," Computers, vol. 13, no. 10, Oct. 2024, Art. no. 264. DOI: https://doi.org/10.3390/computers13100264
K. I. Roumeliotis, N. D. Tselikas, and D. K. Nasiopoulos, "Fake News Detection and Classification: A Comparative Study of Convolutional Neural Networks, Large Language Models, and Natural Language Processing Models," Future Internet, vol. 17, no. 1, Jan. 2025, Art. no. 28. DOI: https://doi.org/10.3390/fi17010028
M. F. Azmi, M. D. A. Kautsar, A. F. Wicaksono, and F. Koto, "IndoSafety: Culturally Grounded Safety for LLMs in Indonesian Languages." arXiv, Jun. 03, 2025. DOI: https://doi.org/10.18653/v1/2025.emnlp-main.465
I. T. Prabowo and P. Purnamasari, "The Influence of Product Reviews and Ratings and Shopee Live on Purchase Decisions through Consumer Trust as an Intervening Variable on Shopee," Review: Journal of Multidisciplinary in Social Sciences, vol. 1, no. 13, pp. 571–580, Dec. 2024. DOI: https://doi.org/10.59422/rjmss.v1i13.707
A. B. H. Krishnan, "Unmasking Falsehoods in Reviews: An Exploration of NLP Techniques." arXiv, Jul. 24, 2023.
M. A. Mohamed, S. D. Ahmed, Y. A. Isse, H. M. Mohamed, F. M. Hassan, and H. A. Assowe, "Detection of Somali-written Fake News and Toxic Messages on the Social Media Using Transformer-based Language Models." arXiv, Mar. 23, 2025.
J. Yi, Z. Xu, T. Huang, and P. Yu, "Challenges and Innovations in LLM-Powered Fake News Detection: A Synthesis of Approaches and Future Directions," in Proceedings of the 2025 2nd International Conference on Generative Artificial Intelligence and Information Security, Hangzhou, China, 2025, pp. 87–93. DOI: https://doi.org/10.1145/3728725.3728739
M. Smith, B. Brown, G. Dozier, and M. King, "Mitigating Attacks on Fake News Detection Systems using Genetic-Based Adversarial Training," in 2021 IEEE Congress on Evolutionary Computation, Kraków, Poland, 2021, pp. 1265–1271. DOI: https://doi.org/10.1109/CEC45853.2021.9504723
K. S. Tarisayi, "Lustre and shadows: unveiling the gaps in South African University plagiarism policies amidst the emergence of AI-generated content," AI and Ethics, vol. 5, no. 1, pp. 245–251, Feb. 2025. DOI: https://doi.org/10.1007/s43681-023-00333-1
X. Tan, J. Gao, and R. Li, "A Simple Structure For Building A Robust Model." arXiv, Jun. 01, 2022. DOI: https://doi.org/10.1007/978-3-031-14903-0_45
N. V. Nguyen, H. Nguyen, Q. Pham, V. Nguyen, S. Ramasamy, and N. Ho, "CompeteSMoE -- Statistically Guaranteed Mixture of Experts Training via Competition." arXiv, May 19, 2025.
X. Feng, D. Song, Y. Chen, Z. Chen, J. Ni, and H. Chen, "Convolutional Transformer based Dual Discriminator Generative Adversarial Networks for Video Anomaly Detection," in Proceedings of the 29th ACM International Conference on Multimedia, Virtual Event, China, 2021, pp. 5546–5554. DOI: https://doi.org/10.1145/3474085.3475693
G. Yin, Y. Pei, S. Farivar, F. Wang, and S. Wang, "Virtual Influencer Marketing: From Social Identification to Parasocial Relationship," in Proceedings of the 58th Hawaii International Conference on System Sciences, Waikoloa, HI, USA, 2025, pp. 2836–2845. DOI: https://doi.org/10.24251/HICSS.2025.342
S. Berry, "Fake Google restaurant reviews and the implications for consumers and restaurants." arXiv, Apr. 27, 2024. DOI: https://doi.org/10.2139/ssrn.4702097
S. Dasgupta and J. Buckley, "A Multi-Embedding Convergence Network on Siamese Architecture for Fake Reviews." arXiv, Jan. 11, 2024.
J. Zeng, Z. Huang, Z. Wu, Z. Chen, and Y. Chen, "FedGR: Cross-platform federated group recommendation system with hypergraph neural networks," Journal of Intelligent Information Systems, vol. 63, no. 1, pp. 227–257, Feb. 2025. DOI: https://doi.org/10.1007/s10844-024-00887-4
K. Coussement and D. F. Benoit, "Interpretable data science for decision making," Decision Support Systems, vol. 150, Nov. 2021, Art. no. 113664. DOI: https://doi.org/10.1016/j.dss.2021.113664
A. Bamdad, A. Owfi, and F. Afghah, "Adaptive Meta-learning-based Adversarial Training for Robust Automatic Modulation Classification." arXiv, Jan. 03, 2025. DOI: https://doi.org/10.1109/ICCWorkshops67674.2025.11162364
Q. Lee, A. Devi, and J. Cutri, "Harnessing the Power of Virtual Reality Experiences as Social Situation of Development to Enrich the Professional Experiences of Early Childhood Pre-Service Teachers," Education Sciences, vol. 15, no. 5, May 2025, Art. no. 635. DOI: https://doi.org/10.3390/educsci15050635
A. Gambetti and Q. Han, "AiGen-FoodReview: A Multimodal Dataset of Machine-Generated Restaurant Reviews and Images on Social Media." arXiv, Jan. 16, 2024. DOI: https://doi.org/10.1609/icwsm.v18i1.31437
K. Yuan, Y. Liu, S. Chandra, and R. Roy, "Retail Market Analysis." arXiv, Jan. 20, 2025.
S. Tufchi, A. Yadav, and T. Ahmed, "AMTCF: an advanced multimodal transformer and ConvNext fusion for contextualized fake news detection in digital landscape," Language Resources and Evaluation, vol. 59, no. 3, pp. 2893–2927, Sep. 2025. DOI: https://doi.org/10.1007/s10579-025-09838-z
N. Khan, T. Nguyen, A. Bermak, and I. Khalil, "CAMME: Adaptive Deepfake Image Detection with Multi-Modal Cross-Attention." arXiv, May 23, 2025.
Downloads
How to Cite
License
Copyright (c) 2025 Sabar Aritonang Rajagukguk, Dedy Sofyan
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain the copyright and grant the journal the right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) after its publication in ETASR with an acknowledgement of its initial publication in this journal.
