Bridging Technology and Law Fake Review Detection Using Naive Bayes Classification and Its Regulatory Implications in E-Commerce

With e-commerce booming, online reviews are an important resource for consumers; however, the area also witnesses the increase in the number of misleading reviews. These deceptive reviews can skew consumer decisions, create an unfair environment for competing businesses, and negatively impact the e-businesses that utilize them as an information source. This study investigates detecting bogus online reviews through a simple machine learning (ML) strategy and assesses its statutory applicability to digital e-businesses. I analysed a dataset containing 4,915 Amazon reviews; applying a rule-based approach to classify "fake" and "genuine" by examining linguistic features of the review (e.g., over-exaggerated phrases, excessive punctuation). It was then classified using a Naive Bayes classifier. According to the findings, 171 reviews were deemed false (3.48% of total) while 4,744 reviews (96.52% of total) were identified as authentic. The model has an overall accuracy of 95.32% for classifying reviews into genuine versus false categories. There is a significant amount of evidence available on how well different machine learning techniques are able to classify whether an online review is true or false, although there are still some limitations when determining whether or not every deceptive review has been detected because of an imbalance in the datasets. In general, however, these results indicate that there may be significant potential to use machine learning technology to help identify and validate online reviews. However, further regulatory oversight and enforcement than currently exists is needed to provide greater consumer protection and transparency in online markets.

Copyright (c) 2026 Journal of Dinda : Data Science, Information Technology, and Data Analytics

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.