Opinion Mining and Sentiment Analysis in an Online

shopping site
Abstract ——Public opinion entails emotions, wishes, attitudes and ideas among individuals
and various social groups in a certain historical stage and social space. This research aims to
employ natural language processing (NLP), sentiment analysis and data mining technologies to
build a public opinion analysis system to serve enterprises' need of online public opinion
detection. By using the data set from an online pear-to-pear (P2P) lending forum with 4148
reviews, we analyzed the data with data visualization techniques.

Keywords-online public opinion; data visualization; sentiment analysis; data mining; opinion
detection

Introduction

In the era of online big data, the continuous development of mobile network technology enables
the application of new media dramatically. Individuals can efficiently spread online information
through social networking sites (SNS). When a growing number of users are interested in social
events, phenomena, comments and supervision, the positive or negative effects of online
information sharing are considerably amplified. However, the vast majority of enterprises do not
have effective means to identify and deal with the negative information which may damage their
reputation and brand value, and information about product market, consumer attitude, industry
development and competitive intelligence, in a timely and accurate manner [3]. At present, many
enterprises adopt traditional manual monitoring methods to approach public opinion. Artificial
monitoring is insufficient to meet the needs of the enterprises. Public opinion related to the focal
enterprise is spread widely on the Internet, but it cannot be easily collected and managed
accurately. This research aims to employ natural language processing (NLP), sentiment analysis
and data mining technologies to build a public opinion analysis system to serve enterprises' need
of online public opinion detection.

Public opinion is the sum of multiple emotions, wishes, attitudes and ideas among individuals
and various social groups in a certain historical stage and social space. Internet public opinion
can be developed from online speech and is a specific form of public opinion. Whether Internet
speech can develop into Internet public opinion largely depends on whether the topic of Internet
speech has strong activeness and sensitivity. Online public opinion is diversified, spread quickly
and interactive, which has the incomparable advantages over traditional media. Online public
opinion has positive and healthy aspects, as well as negative and decadent aspects.

The enterprise public opinion is limited to the public opinion content related to a certain
enterprise subject. It is a collection of a variety of emotions, attitudes and opinions expressed
through the Internet by the vast number of Internet users in a specific period for certain product
production or service provided by a specific enterprise. With greater use of online media,
enterprises pay more and more attention to the impact of their online public opinion on their
corporate image. Network blog, SNS (brand forum, WeChat, QQ, etc.), and e-commerce
platform are the main channels of online public opinion communication. The enterprise networks
public opinion has the general characteristics of network public opinion: the virtual space of
public opinion information, the real-time nature of public opinion communication, the
interactivity of opinion published, disseminated and fed back by online participants, and the
openness of online opinion communication.

The network public opinion of enterprises often spreads and breaks out in a large scale in a short
period of time due to an emergency, and its impact will be immediately transmitted to the
production and sales performance of enterprises. Therefore, the timeliness of monitoring and
management of network public opinion of enterprises is relatively high. However, the
government's management of social public opinion involves a larger scope, the time for correct
guidance of online public opinion is relatively long, and the influence of public opinion is more
average.

The public opinion management of the enterprise has strong target, mainly for the relevant
groups of the enterprise, including customers, partners, main competitors, potential target
customers, etc. The government level social public opinion management aims at the whole
country's all-round public opinion management, which is broader and less targeted compared
with enterprises. The management of public opinion in enterprises tends to be actively used and
exerted, while the management of public opinion at the government level tends to be supervised
and prevented. Specifically, corporate public opinion is generally to collect, summarize, analyze
and further spread the information of corporate brand reputation, competitor dynamics, industry
status and hot events, to achieve the purpose of commercial profit [11]. While the government
level social public opinion management focuses on the use of the network to timely discover the
information that has an adverse impact on national interests and timely supervise and handle, so
as to minimize the adverse impact on society.
Literature Survey

 IEEE Papers

Sr. No. Paper / Publication Author Year

1. Mining topical relations between Prajkta Akre;Harshali Patil;Anand 2017

opinion word and opinion target. Khandare;Mohammad Atique
2. Opinion mining on author's citation S. Anupkant;P.V.M. Seravana 2017
characteristics of scientific publications Kumar;Nayani Sateesh;D. Bhanu
Mahesh
3. Aspect Level Opinion Mining for Hotel Cho Cho Hnin;Naw Naw;Aung Win 2018
Reviews in Myanmar Language.
4. Product Quality Assessment using Fatemeh Hosseinzadeh 2019
Opinion Mining in Persian Online Bendarkheili;Rezvan
Shopping MohammadiBaghmolaei;Ali
Ahmadi
5. An Enhanced Architecture for Feature A. Angelpreethi;S. Britto Ramesh 2017
Based Opinion Mining from Product Kumar
Reviews.
6. Evaluating Performance of Machine Prerna Mishra;Ranjana 2018
Learning Techniques used in Opinion Rajnish;Pankaj Kumar
Mining
7. Review on Natural Language Yasir Ali Solangi;Zulfiqar Ali 2018
Processing (NLP) and Its Toolkits for Solangi;Samreen Aarain;Amna
Opinion Mining and Sentiment Abro;Ghulam Ali Mallah;Asadullah
Analysis. Shah
8. Yasir Ali Solangi;Zulfiqar Ali Pawan Kumar Verma;Shalini 2017
Solangi;Samreen Aarain;Amna Agarwal;Mohd Aamir Khan
Abro;Ghulam Ali Mallah;Asadullah
Shah
9. News Comments Modeling for Opinion Lamine Faty;Marie Ndiaye;Edouard 2020
Mining: The Case of Senegalese Online Ngor Sarr;Ousmane Sall;Sény
Press Ndiaye Mbaye;Tony Tona
Landu;Babiga Birregah;Mamadou
Bousso
10 Twitter Opinion Mining and Boosting R Geetha;Pasupuleti Rekha;S 2018
Using Sentiment Analysis Karthika
Problem Definition

To Develop natural language processing (NLP), sentiment analysis and data mining technologies
to build a public opinion analysis system to serve enterprises' need of online public opinion
detection.

Purpose

Understanding people’s emotions is essential for businesses since customers express their
thoughts and feelings more openly than ever before. Automatically analyzing customer feedback,
such as opinions in survey responses and social media conversations, allows brands to listen
attentively to their customers, and tailor products and services to meet their needs.

Goals and Objectives

The Goals and Objectives of the given system are as follows:

 To build a public opinion analysis system for business to understand customer’s need.
 To analyze the data using various visualization techniques (Chart, Map, Graphs)
 To develop a system that helpful for businesses to pay more attention to the impact of
their public online on their corporate image and take various decisions.

Features

 Pre-processing is done to improve accuracy of Data.

 Fast and High Accuracy.
 AI and ML is used to improve the accuracy of System.
 Allows much more detailed sentiment analysis on each entity
 Listen to your user/customers in real time and make data-based decisions on the go.
Proposed Architecture

Fig – Proposed Architecture

Methodology

Opinion Mining also called sentiment analysis is a process of finding user's opinion towards a
topic or a product. Opinion mining concludes whether user's view is positive, negative, or neutral
about product, topic, event etc. Opinion mining and summarization process involve three main
steps, first is Opinion Retrieval, Opinion Classification and Opinion Summarization. Review
Text is retrieved from review websites. Opinion text in blog, reviews, comments etc. contains
subjective information about topic. Reviews classified as positive or negative review. Opinion
summary is generated based on features opinion sentences by considering frequent features about
a topic.

Software Required

• Programming language – Python (For Developing API)

• Libraries – NumPy, NLTK,
 • Tool – PyCharm, FileZilla, PuTTY (SSH client), Postmen
• Database – MySQL
• Web – HTML, CSS, JavaScript, Bootstrap, PHP

Area of Project

Machine Learning, NLP, Web Application

Advantages

 Upselling opportunities
 Adaptive customer service
 Reduce customer churn
 Tracking overall customer satisfaction
 Detect changes in customer opinion

Limitations

 Detection of spam and fake reviews.

 Domain-independence
 Natural language processing overheads

Conclusion

Sentiment analysis can be applied to countless aspects of business, from brand monitoring and
product analytics, to customer service and market research. By incorporating it into their existing
systems and analytics, leading brands (not to mention entire cities) are able to work faster, with
more accuracy, toward more useful ends.

Sentiment analysis has moved beyond merely an interesting, high-tech whim, and will soon
become an indispensable tool for all companies of the modern age. Ultimately, sentiment
analysis enables us to glean new insights, better understand our customers, and empower our
own teams more effectively so that they do better and more productive work.

References

1. Product Quality Assessment using Opinion Mining in Persian Online Shopping Fatemeh
HosseinzadehBendarkheili;Rezvan MohammadiBaghmolaei;Ali Ahmadi 2019 27th Iranian Conference on Electrical
Engineering (ICEE) Year: 2019 | Conference Paper | Publisher: IEEE

2. News Comments Modeling for Opinion Mining: The Case of Senegalese Online Press Lamine Faty;Marie
Ndiaye;Edouard Ngor Sarr;Ousmane Sall;Sény Ndiaye Mbaye;Tony Tona Landu;Babiga Birregah;Mamadou Bousso
2020 International Conference on Advances in Computing and Communication Engineering (ICACCE) Year: 2020 |
Conference Paper | Publisher: IEEE

3. Chinese Explanatory Opinion Relationship Recognition Based on Improved Target Attention Mechanism Xue
Cao;Chenghao Zhu;Chengguo Lv 2020 3rd International Conference on Advanced Electronic Materials, Computers
and Software Engineering (AEMCSE) Year: 2020 | Conference Paper | Publisher: IEEE

4. Opinion Mining Using Live Twitter Data Andleeb Aslam;Usman Qamar;Reda Ayesha Khan;Pakizah Saqib;Aleena
Ahmad;Aiman Qadeer 2019 IEEE International Conference on Computational Science and Engineering (CSE) and
IEEE International Conference on Embedded and Ubiquitous Computing (EUC) Year: 2019 | Conference Paper |
Publisher: IEEE

5. An Opinion Mining for Indian Premier League Using Machine Learning Techniques Arti;Kamanksha Prasad
Dubey;Sanjy Agrawal 2019 4th International Conference on Internet of Things: Smart Innovation and Usages (IoT-
SIU) Year: 2019 | Conference Paper | Publisher: IEEE

6. The Essential of Sentiment Analysis and Opinion Mining in Social Media : Introduction and Survey of the Recent
Approaches and Techniques Hoong-Cheng Soong;Norazira Binti A Jalil;Ramesh Kumar Ayyasamy;Rehan Akbar 2019
IEEE 9th Symposium on Computer Applications & Industrial Electronics (ISCAIE)
7. Year: 2019 | Conference Paper | Publisher: IEEE

8. Mining User Opinion Influences on Twitter Social Network: Find that Friend who Leads your Opinion Using Bayesian
Method and a New Emotional PageRank Algorithm Armielle Noulapeu Ngaffo;Walid El Ayeb;Zied Choukair 2019
15th International Wireless Communications & Mobile Computing Conference (IWCMC) Year: 2019 | Conference
Paper | Publisher: IEEE

9. Survey on Opinion Leader in Social Network using Data Mining Rahul Katarya;Diksha Gautam 2019 5th
International Conference on Advanced Computing & Communication Systems (ICACCS) Year: 2019 | Conference
Paper | Publisher: IEEE

10. Public Opinion Detection in an Online Lending Forum: Sentiment Analysis and Data Visualization Ge Zhan;Ming
Wang;Meiyi Zhan 2020 IEEE 5th International Conference on Cloud Computing and Big Data Analytics (ICCCBDA)
Year: 2020 | Conference Paper | Publisher: IEEE

11. https://www.sciencedirect.com/topics/computer-science/opinion-mining

12. https://www.brandseye.com/news/what-is-opinion-mining-next-level-sentiment-analytics/

13. https://searchbusinessanalytics.techtarget.com/definition/opinion-mining-sentiment-mining