Title: Detection of fraudulent and malicious websites by analysing user reviews for online shopping websites

Authors: Asha S. Manek; P. Deepa Shenoy; M. Chandra Mohan; K.R. Venugopal

Addresses: Department of Computer Science and Engineering, Jawaharlal Nehru Technological University, Hyderabad, India ' Department of Computer Science and Engineering, University Visvesvaraya College of Engineering, Bangalore University, Bangalore 560 001 India ' Department of Computer Science and Engineering, Jawaharlal Nehru Technological University, Hyderabad, India ' Department of Computer Science and Engineering, University Visvesvaraya College of Engineering, Bangalore University, Bangalore 560 001 India

Abstract: Recently, the web has become a crucial worldwide platform for online shopping. People go online to sell and buy products, use online banking facilities and even give opinions about their online shopping experience. People with malicious intent may be involved in any online transaction with a fraudulent e-business give fake positive reviews that actually does not exist to promote or degrade the product. User reviews are extremely essential for decision making and at the same time cannot be reliable. In this paper, we propose a novel method Bayesian logistic regression classifier (BLRFier) that detects fraudulent and malicious websites by analysing user reviews for online shopping websites. We have built our own dataset by crawling reviews of benign and malicious e-shopping websites to apply supervised learning techniques. Experimental evaluation of BLRFier model achieved 100% accuracy signifying the effectiveness of this approach for real-life deployment.

Keywords: fake reviews; malicious websites; supervised learning; sentiment analysis; Bayesian logistic regression; fraud detection; user reviews; online shopping websites.

DOI: 10.1504/IJKWI.2016.078712

International Journal of Knowledge and Web Intelligence, 2016 Vol.5 No.3, pp.171 - 189

Received: 09 Oct 2015
Accepted: 02 Dec 2015

Published online: 26 Aug 2016 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article