International Journal of Trend in Scientific Research and Development (IJTSRD)

Volume: 3 | Issue: 3 | Mar-Apr 2019 Available Online: www.ijtsrd.com e-ISSN: 2456 - 6470

Fake Product Review Monitoring System

Piyush Jain, Karan Chheda, Mihir Jain, Prachiti Lade
Department of Computer Engineering, Shah and Anchor Kutchhi Engineering College,
Chembur, Mumbai, Maharashtra, India

How to cite this paper: Piyush Jain | ABSTRACT

Karan Chheda | Mihir Jain | Prachiti In the current scenario, the data on the web is growing exponentially. Social
Lade "Fake Product Review media is generating a large amount of data such as reviews, comments, and
Monitoring System" Published in customer’s opinions on a daily basis. This huge amount of user generated data is
International Journal of Trend in worthless unless some mining operations are applied to it. As there are a number
Scientific Research and Development of fake reviews so opinion mining technique should incorporate Spam detection
(ijtsrd), ISSN: 2456-6470, Volume-3 | to produce a genuine opinion. Nowadays, there are a number of people using
Issue-3, April 2019, pp. 105-107. social media opinions to create their call on shopping for product or service.
http://www.ijtsrd.co Opinion Spam detection is an exhausting and hard problem as there are many
m/papers/ijtsrd216 faux or fake reviews that have been created by organizations or by the people for
44.pdf various purposes. They write fake reviews to mislead readers or automated
detection system by promoting or demoting target products to promote them or
Copyright © 2019 by to degrade their reputations. The proposed technique includes Ontology, Geo
author(s) and location and IP address tracking, Spam words Dictionary using Naïve Bayes,
International Journal Brand only review detection and tracking account used.
IJTSRD21644
of Trend in Scientific
Research and Development Journal. This
is an Open Access article distributed
under the terms of the Creative
Commons Attribution License (CC BY
4.0)
(http://creativeco
mmons.org/license
s/by/4.0)
1. INTRODUCTION 2. RELATED WORK
One of the very rapid growth area is ecommerce. Generally 2.1 Opinion Mining Using Ontological Spam Detection
e-commerce provide facility for customers to write reviews Duhan & Mittal proposed a paper “Opinion Mining Using
related with its service. The existence of these reviews can Ontological Spam Detection” which will help us to find out
be used as a source of information. For examples, companies fake reviews by using Naïve Bayes as algorithm. To find out
can use it to make design decisions of their products or fake review in the website this “Fake Product Review
services but unfortunately, the importance of the review is Monitoring System” system is introduced. This system will
misused by certain parties who tried to create fake reviews, find out fake reviews made by the customers and it will
both aimed at raising the popularity or to discredit the block the users. To find out the review is fake or genuine, we
product. They share their thoughts on internet. will use some classification such as

Before purchasing anything, it is a normal human behavior  Tracking IP address of the user to detect if the reviews
to do a survey on that product. Based on reviews, customers are from a Spammer. If multiple reviews are from the
can compare different brands and can finalize a product of same IP address then the Reviews are considered Spam.
their interest. These online reviews can change the opinion  Using Account Used to check whether the reviews are
of a customer about the product. If these reviews are true, done using the same account.
then this can help the users to select proper product that  Brand only Review detection i.e. whether the reviews
satisfy their requirements. On the other hand, if the reviews are on only Brand not the product. It’s not helpful to
are manipulated or not true then this can mislead user. This consider only the Brand value to judge a product.
boosts us to develop a system which detect fake reviews for  Using Negative Dictionary i.e. the negative words are
a product by using the text and rating property from a identified in the review. If there are more than five
review. The honesty value and measure of a fake review will Negative Words then the review is a Spam.
be measured by utilizing the data mining techniques.
For instance, a user has posted a Review: “This product
An algorithm could be used to track customer reviews, is not good, the design is bad, quality is worst and it is
through mining topics and sentiment orientation from online worthless to buy.” Here, this sentence consists of 4-5
customer reviews and will also blocked the fake reviews. negative words. So, the system will check the count of
negative words, if the count exceeds, then it will be
considered as spam review. Therefore Negative Word

@ IJTSRD | Unique Paper ID - IJTSRD21644 | Volume – 3 | Issue – 3 | Mar-Apr 2019 Page: 105
 International Journal of Trend in Scientific Research and Development (IJTSRD) @ www.ijtsrd.com eISSN: 2456-6470
Dictionary will be used with customized Senti strength  Admin Login: - Admin login to the system using his
algorithm. According to this approach, probability of admin ID and password.
given review to be Spam is more so it will be considered  Add product: - Admin will add product to the system.
a Spam.  Delete Review: - Admin will remove the review which
 Using Ontology: For instance, if the review posted on a tracked by the system as fake.
product is not about that product but talking about  User Login: - User will login to the system using his user
something else then ontology is used to identify and ID and password.
classify such reviews as spam.  View product: - User will view product.
 Post Review: - User can post review about the product.
If Class: Toshiba
3. PROBLEM DEFINITION AND OBJECTIVES
Context: Laptop Review: Dell is not so good.  PROBLEM DEFINITION:-
In recent years, online reviews have been playing an
Here User is Posting Reviews about Laptop that comes under important role in making purchase decisions. This is
the class Toshiba. But his Review contains Dell Keyword. In because, these reviews can provide customers with large
order to identify this Review as Spam we are going to use amounts of useful information about the goods or service.
Ontology. However, to promote factitiously or lower the quality of the
products or services, spammers may forge and produce fake
This system uses data mining methodology and Opinion reviews. Due to such behavior of the spammers, customers
mining technology. This system helps the user to find out would be mislead and make wrong decisions. Thus detecting
correct review of the product, will also help the user to fake (spam) reviews is a significant problem. Opinion
detect fake review and makes them to block the fake reviews spamming refers to the use of excessive and illicit methods,
automatically. such as creating a large volume of fake reviews, in order to
generate biased positive or negative opinions for a target
2.2 Fake Product Review Monitoring and Removal for product or service with the intention of promoting or
Genuine Online Product Reviews Using Opinion demoting it, respectively. The reviews created for this
Mining. purpose are known as fake, spam or bogus reviews, and the
Kohli, Mishra & Gupta proposed a paper “Fake Product authors responsible for composing such deceptive content
Review Monitoring and Removal for Genuine Online Product are known as fake or spam reviewers.
Reviews Using Opinion Mining” which help us in detecting
the fake reviews and track down the user. As most of the  OBJECTIVES :-
people require review about a product before spending their The identified challenges motivate to bring up a solution to
money on the product. So people come across various all the problems stated in the above problem statement
reviews in the website but these reviews are genuine or fake section. Following are the objectives of the proposed
is not identified by the user. In some review websites some approach and this thesis work:
good reviews are added by the product company people  To implement different algorithm to get better Spam
itself in order to make product famous this people belong to Detection i.e.; IP Address, Account used, Negative Word
Social Media Optimization team. Dictionary using Senti-strength, Ontology.
 Graphical representation of work.
They give good reviews for many different products  To deals with 6 different types of Spam Reviews.
manufactured by their own firm. User will not be able to find  To presents Opinion Mining on Spam Filtered Data.
out whether the review is genuine or fake. To find out fake  To implement Ontology in Spam Detection
review in the website this “Fake Product Review Monitoring  To present an algorithm that does Opinion Mining with
and Removal for Genuine Online Product Reviews Using Spam Detection.
Opinion Mining” system is introduced. This system will find
out fake reviews made by the social media optimization team 4. FLOWCHART
by identifying the IP address. User will login to the system
using his user id and password and will view various
products and will give review about the product. And the
user will get genuine reviews about product. And while
reviewing he needs to enter the email id from which he is
reviewing and it would be verified. If he writes a fake review
then his id will be blocked bot allowing him to share his
opinions again.

System works as follows:-

 Admin will add products to the system.
 User need to enter their email id and OTP no to enter the
system
5. SUMMARY
 User once access the system, user can view product and
They are various ways to detect Spam Reviews in order to
can post review about the product.
the Opinion mining to be more accurate and useful have
 For posting reviews, the user’s id will be verified.
been studied. A detailed discussion about the existing
 And admin will also block the email id of the user if
techniques, to find out the whether the review is spam or not
reviews are spammed.
is presented. Other Techniques are incorporated like IP
 Admin will delete the review which is fake.
Address Tracking and Ontology to detect Spam Reviews in
order to get more accurate results from Opinion mining.

@ IJTSRD | Unique Paper ID - IJTSRD21644 | Volume – 3 | Issue – 3 | Mar-Apr 2019 Page: 106
 International Journal of Trend in Scientific Research and Development (IJTSRD) @ www.ijtsrd.com eISSN: 2456-6470
After detecting the spam reviews from the existing Dataset, a
new Dataset is created which doesn’t contain spam reviews
and then opinion mining is performed on the new Spam REFERENCES
Filtered Dataset. At last a new algorithm is proposed that [1] Rajashree S. Jadhav, Prof. Deipali V. Gore, "A New
detects spam reviews more precisely and performs opinion Approach for Identifying Manipulated Online Reviews
mining using spam filtered data. using Decision Tree ". (IJCSIT) International Journal of
Computer Science and Information Technologies, Vol. 5
Working (2), pp 1447-1450, 2014
[2] Long- Sheng Chen, Jui-Yu Lin, “A study on Review
1. User will be allowed to review only if he is logged into
Manipulation Classification using Decision Tree", Kuala
our online portal.
Lumpur, Malaysia, pp 3-5, IEEE conference publication,
2. After logging in user will be allowed to review for the 2013.
product.
[3] Benjamin Snyder and Regina Brazil, “Multiple Aspect
3. Once the user enters the review, the reviews will be
ranking using the Good Grief Algorithm “Computer
processed and analyzed for spam on following
Science and Artificial Intelligence Laboratory
conditions:
Massachusetts Institute of Technology2007.
a) Does the review entered by the user contain any
link which redirects them to other product page for [4] Ivan Tetovo, “A Joint Model of Text and Aspect Ratings
brand promotion? for Sentiment Summarization “Ivan Department of
Computer Science University of Illinois at Urbana, 2011
b) Analyzing whether multiple review have come from
the same user. [5] N. Jindal and B. Liu, “Analyzing and detecting review
c) Analyze whether same email account or same ip- spam,” International Conference on Web Search and
address are used for multiple reviews on same Data Mining, 2007, pp. 547-552.
product. [6] N. Jindal and B. Liu, “Opinion spam and analysis,”
d) Analyze the reviews or ratings to detect whether International Conference on Web Search and Data
reviews are spam or not. Mining, 2008, pp. 219-230.
4. If the review posted by the user satisfies any of the above
specified conditions then it will be considered as spam/fake
reviews.
5. Once the review is detected as spam review or fake
review, then user account will be blocked and review will be
reported to the administrator.

@ IJTSRD | Unique Paper ID - IJTSRD21644 | Volume – 3 | Issue – 3 | Mar-Apr 2019 Page: 107