An overview of soft computing techniques on Review Spam Detection

Abstract

Individuals that work with web services commonly focus on internet-based feedback to make decisions about shopping via online, booking an apartment, renting banquette halls etc., because people believe that making a decision which is based on the views of others, can contribute in making the right decision about their purchases. While publishing negative reviews or comments, brings financial benefit to manufacturers via online review websites and also increases the review spam happenings. Sentimental misinformation in feedback has therefore become a significant challenge for people to make decisions about their purchase and thus decrease the credibility of review on websites. The malicious spam identification of views is therefore an important activity in the fields of machine learning (NLP- Natural Language Processing). The conventional bag-of-words model is used to recognize the text analysis characteristics of most of the current research on review spam detection and to apply traditional machine learning models such as Support Vector Machines, Random forest, Decision tree and Naive Bayes as classifiers. With increase in the effective applications of Soft computing, the spam detection framework has achieved improved efficiency over traditional machine learning and works well with unique models and provides better solutions to complicated real-life scenarios. This paper discusses the overview of soft computing techniques in spam detection.