Machine Learning Algorithms for Detecting Fake Reviews – A Study

Fake reviews are very popular today as every internet user is depending on the reviews to draw a conclusion on a particular product or organization or hotels in their respective domains. Fake reviews are the one and only thing that is affecting the present world businesses and all other things. As development of web 2.0 it's seen that customers rely more on product ratings and reviews given by customers for product information. Usually, a more positive review attracts more customers and with intern is responsible for the growth of that product. Nowadays people take advantage of this fact and give false reviews in order to mislead the customers affecting company, market efficiency and merchants. The issue which plays the major role in the field of fake review detection is that the fake review groups are adapting to the present users’ expectations and creating the fake reviews which are not able to be detected by the humans. So, we are trying to solve it using some technologies like machine learning to “Fight the machines with machines” to detect the fake reviews using these technologies. And there are many methods in which these fake reviews can be detects and taken care of using many types of machine learning methods. This article aims to achieve understanding of fake reviews detection in various domains like e-commerce and hotel websites etc. And we also work to identify the relation between the fake review and the reason behind creating them. By performing a systematic literature review we can identify and understand about the main approaches that have been previously studied and made accessible to recognize fake reviews and their applications in different situations. And also, absence of high- quality datasets limit the research of fake reviews. We have also mentioned existing research problems in data acquisition and identifying method design to help in future research on fake review detection.