Machine Learning Methods for Fake Review Detection

Abstract

This paper investigates the problem of automatically detecting fake reviews of products and 
services using machine learning methods. The rise of online commerce has made it necessary to develop 
reliable tools for identifying deceptive reviews that distort consumer perceptions and damage businesses' 
reputations. The study explores text analysis and basic classification methods using algorithms such as 
Naive Bayes and Random Forest, employing TF-IDF text representation. The paper presents a comparative 
analysis of these methods and discusses their effectiveness in terms of accuracy, precision, recall, and F1
score.