Abstract: Online reviews in modern businesses and e-commerce platforms are influential and this is an important factor for customers. Potential buyers heavily rely on user’s feedback when deciding whether or not to purchase products online. Unfortunately, there are some unscrupulous individuals and organizations that attempt to manipulate these reviews to suit their interests. As a result, fake reviews increase and mislead customers/buyers.
To address this issue, a proposed solution involves the implementation of an e-commerce platform that utilizes an NLP algorithm to analyze the features and sentiments of reviews and a data science algorithm to classify any fraudulent text. The proposed system contains a review detection model within an e-commerce application, where customers may register, view products, and make purchases. As customers buy products, they post reviews and ratings. Using these inputs, the model is able to classify given reviews as false or true using KNN classification algorithm.
Keywords: Machine Learning, NLP, Fake Reviews, Classification Algorithm, KNN, Lesk
|
DOI:
10.17148/IJARCCE.2023.124120
