International Journal of Engineering Trends and Applications (IJETA) – Volume 11 Issue 6 Nov - Dec 2024 ISSN: 2393-9516 www.ijetajournal.org Page 25 Consumer Spam Review Detection With Linguistic Method Jyoti G.Biradar Department of Computer Science, Rani Channamma University, Belagavi, Karnataka – India ABSTRACT With the emergence of popularity of E-commerce websites they also provide an option to share customer’s opinion or review about their products. Thus, before purchasing any product (or) service it helps the customer to take decisions and for the vendors to promote their product and increase its sales. But, all the reviews posted by the users (or) customers need not to be written with true intention. User might have posted his review to promote (or) demote the product, such users or customers are considered as spammers. Detecting review spam is a challenging task due to the openness of writing product review on company’s websites Thus, need to propose a new effective approach to classify each and every review as spam or non-spam. In this paper, the work is proposed for detecting spam/fake reviews on products using Term Frequency-Inverse Document Frequency (TF-IDF) weightage. Keywords: Reviews, spam detection, fake reviews, spammers, Term Frequency-Inverse Document Frequency. I. INTRODUCTION It has become a habit for purchasing any product first they go through the reviews regarding the products, company or service etc. Sharing reviews through websites has become an important source of customer’s opinions. Therefore, if number of positive responses are more than the customers likes to buy the product. If the reviews are in terms of negative then customers move towards other products. Thus good reviews affect the economic condition and frame of organizations. The other part of these framework spammers tries to generate fake reviews for promotions or demotions of brands and mislead the customers. The main disadvantage here is there is no filtering or control on writing reviews as anyone can login and write their reviews. These reviews become an important source of data for buyers to decide and purchase the products. Thus spammers have chance to participate and affect the reputation of the products of the company. The term spam describes unwanted message delivered to a large number of users from any company or website. Spam can be in many forms such as unwanted political mails, education, health care, personal, finance, computers, automotive, adult content and more. With the emergence of popularity of E-commerce websites they also provide an option to share customer’s opinion or review about their products. Spammers can be defined as the type of malicious users who damages the information presented by legitimate users and also in turn fetches risk to the security and privacy of social networks. Spam is identified with following scenario: • Unwanted junk mails are arrived in the mail box • Generates burden for communications service providers and business to filter email. • Phishing with the information by tricking into different links or entering details with good offers and promotions. • Phishes: Behavior of the users matches the normal user to steal personal information of other legal users. • Fake Users: Users who impersonate the profiles of genuine users to send spam content to the friends of that user or other users in the network. • Promoters: are the ones who send malicious links of advertisements or other promotional links to others so as to obtain their personal information. Spam reviews can be categorized into two types. In first type of reviews, spammers try to mislead readers by expressing undeserving positive opinions for some selected products for promotions or by writing negative reviews for damaging the name of the products. The second type of reviews includes only advertisements. Usually spammers focus on few goods and their service that users of computer are likely to be interested and they select the grey or black goods from market. In other way, spam can be stated as illegal, not only because of advertising the goods, but of the goods and services offered through advertisements are also illegal. From either sides business or company, reviews have following advantages: First, increase in rates of sales and conversion. Secondly, understand the feelings of customers about their brands, likes and dislikes about the products, improvements in shopping experience. Third, monitor and improve service offered by the company. Fourth, increase the traffic to website. Fifth, increase the loyalty of the customers. Furthermore, opinions are important for management of reputation and brand perception of product. Detecting review spam is a challenging task due to the openness of writing product review on company’s websites, there is no large-scale ground truth labeled dataset available. Spammers usually pose by different names of users which makes harder to eradicate spam reviews completely. Spam reviews looks perfectly normal until comparing with other reviews of the same products to identify review comments not consistent with the latter. The efforts of additional comparisons by the users make the detection task tedious and non-trivial.In this paper, Review spam identification is an important component to identify review spam, the data set consists of about 60k reviews. We need to provide real and trustful review mining results. In this paper, we introduce our review spam identification based on calculating term frequencies and using similarity measure. RESEARCH ARTICLE OPEN ACCESS