• Ajeet Singh
• Aashish Panwar
• Dr.Vibha Srivastava

With the continual evolution of E-commerce platforms, online evaluations are increasingly seen as a critical aspect in establishing and maintaining a positive reputation. Furthermore, they play an important role in the decision-making process for end users. A positive review for a specific object typically draws more customers and leads to a significant rise in sales. Deceptive or phony evaluations are now purposefully generated in order to build a virtual reputation and attract potential clients. As a result, detecting false reviews is an active and ongoing research topic. Identifying phony reviews is dependent not only on the essential elements of the reviews, but also on the reviewers' behaviour. This research provides a machine learning method for detecting false reviews. In addition to the review features extraction approach, this research employs different features engineering techniques to extract diverse reviewer behaviours. The study examines the performance of machine learning classifiers; KNN, Naive Bayes, and Logistic Regression using a genuine Yelp dataset of restaurant reviews. In terms of accuracy, the results show that Logistic Regression surpasses the other classifiers. The results demonstrate that the system is better at determining whether a review is fake or genuine.

Machine learning, fake, reviews, Naive Bayes, KNN, Logistic Regression

"Partially Supervised Learning Group Detection", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.10, Issue 4, page no.m225-m229, April-2023, Available :http://www.jetir.org/papers/JETIR2304C25.pdf

"Partially Supervised Learning Group Detection", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.10, Issue 4, page no. ppm225-m229, April-2023, Available at : http://www.jetir.org/papers/JETIR2304C25.pdf