You are on page 1of 3

Sentiment Analysis for Promotional Campaigns

1st Sameer Mulani 2nd Nikhat Pathan


Department of Computer Engineering Department of Computer Engineering
Anjuman-I-Islam’s Kalsekar Technical Campus Anjuman-I-Islam’s Kalsekar Technical Campus
Mumbai, India Mumbai, India

3rd Saffana Siddiqui 4th Zeeshan Ansari


Department of Computer Engineering Department of Computer Engineering
Anjuman-I-Islam’s Kalsekar Technical Campus Anjuman-I-Islam’s Kalsekar Technical Campus
Mumbai, India Mumbai, India

Abstract—Sentiment Analysis is a way to know one’s opinion data on the social media includes the opinionsof the user.
on a particular topic. The study helps in determine whether the Twitter is very much popular now a days in various areas
subject in analysis has a positive impact or a negative impact. e.g. the election exit polls are deduced by analyzing trends on
A promotional campaign of a product or service helps to reach
to audience and to make them aware of the product or service. twitter during pre-election period. Social media contentsspread
Our study uses natural language processing to determine the swiftly while users communicate and share them freely within
sentiments in the content from social media post. The social the site. Among social media sites, Twitter has very high
media sites are inherent with information like users opinions on communicability of information on social issues, disasters and
a specific product or activities. Also, it is useful for gathering accidents. In social media sites, which are directly used in real
an overview of users opinions on a large-scale on certain
topics. The potential to extract perceptions from social sites are life and enable the share of numerous pieces of information
globally embraced by various organisations. It plays vital role in in real time, contents created by users become major infor-
different professions and marketing. The objective of project is mation. Contents that contain information or sentiments with
to provide most recent tweets that can help in examining users photographs, short writings, andeven only one word such as
opinions which can be further used for predictions in marketing, a hash tag are shared. As such, the use of social media sites
campaigning , entertainment industries.
Index Terms—sentiment analysis, opinion mining, machine has already penetrated into life to play the role of important
learning, python, text mining, stemming, tokenizing, text clas- contents in diverse fields. We use socialmedia contents as
sification. a sensor for event detection. The social media contents that
are continuouslygenerated in real time show social issues and
I. I NTRODUCTION the daily lives of individuals. Due to the rapid increase in
In today’s era, people tend to speak and give their opinions the textual data in social media platforms, there is a need
on every topic on the internet. Hence, internet is now a to analyze the concept of expressing sentiments. Business
huge and easy source of opinions. Thus, the study of these owners and advertising companies often employ sentiment
opinions can be extremely effective in performance analysis. analysis to discover new business strategies and advertising
As campaigns are seen to take a great height from the past campaign.The proposed paper fetches the tweets from Twitter
few years, there exists a need to know the audience reactions. based on keywords from the users. It will then process it
Here, the sentiment analysis of the opinions of these audiences using various techniques like stemming, stop-word removal
may help in measuring the success of campaigns. Campaigns etc and then finally it generates the result in statistical form
prove to be important part in reaching the audience, hence like percentage etc.
it is necessary to monitor the success of the campaigns in
further planning. This would benefit many business in order II. M ETHODOLOGY
to promote their product which would result their sales to rise The project is split into 2 phases, viz, the literature study
and also many social workers regarding their social campaigns. and the development phase. The literature study involves study
With the ease of access to the internet, the social media users of various twitter API, natural language processing techniques,
share their views on their social accounts. In contrast to other text classification and machine learning algorithms. The study
social platforms, twitter is found to be a reliable source. The of current systems in the market and their drawbacks is also
opinions from the posts is extracted using machine learning a part of this phase. In prior to the development phase,
techniques and natural language processing. the functional and non-functional requirements are taken into
In today’s world Social media plays a key role in recording considerations. The implementation logic and flow is built and
our day to day life responses. TheInternet has provided a the application design is discussed. The Development Phase
platform for people to express their views, emotions and involves use of Python 3, twitter API and data warehousing
sentimentstowards products, people and life in general. The techniques in terms of implementation.
III. L ITERATURE R EVIEW A. Data Caching
A. Natural Language Processing
Natural Language Processing refers to AI method of com- Usually, a tweet consists of one or more than one keyword
municating with an intelligent systems using a natural lan- or hashtags. Also, particular topic of campaign can be a
guage such as English. Processing of Natural Language is collection of hashtags. There exists a high chance that a
required when you want an intelligent system like robot to hashtag can be used for many tweets related to different
perform as per your instructions, when you want to hear campaigns. For this we have implemented a Data caching
decision from a dialogue based clinical expert system, etc. module, it saves the fetched tweets in the database and also
Steps in NLP: it sentiment after it is processed. This reduces the redundancy
• Lexical Analysis of the system and hence only the new tweets are required. As
• Syntactic Analysis there are huge amount of tweets related to trending hashtags,
• Semantic Analysis this module would not only speed up system, but also make
• Disclosure Integration the system provide up-to-date results.
• Pragmatic Analysis

B. Twitter
Twitter is an American online news and social networking
service on which users post and interact with messages known
as ”tweets”. Tweets were originally restricted to 140 charac- B. Text pre-processing
ters, but on November 7, 2017, this limit was doubled for all
languages except Chinese, Japanese, and Korean. Registered
users can post, like, and retweet tweets, but unregistered users
This module cleans the content, pre-processes the data
can only read them.
and transform it into machine suitable form. This involves
Twitter is used to connect people with the same interests.
Tokenizing, Stop-word Removal and Stemming.
As the Twitter homepage suggests, the social platform can be
used to, “Connect with your friends — and other fascinating Tokenizing phase breaks the content into arrays of words.
people. Get in-the-moment updates on the things that interest This makes it easy to detect the words for Parts of Speech.
you.” This makes twitter to shine in comparison to other social
platforms when it comes to discussion. As the Parts of Speech can be detected, stop-words such as
Users can group posts together by topic or type by use of articles and interjection can be removed from the sentences.
hashtags – words or phrases prefixed with a ”” sign. Similarly, Stemming is an important module in the system as the word
the ”@” sign followed by a username is used for mentioning cannot be classified easily. Words should be transformed into
or replying to other users. their normal form and their suffix should be replaced with
For sentiment analysis, twitter proves to be the best fit as appropriate suffix.
it focuses on similar interests. We use hashtags to search for
post on our required topic.

C. Campaigns
A series of advertisements using various marketing tools
that share the same message and ideas to promote a business
C. Sentiment Analysis
or event to a target audience.
Phone calls, door-to-door visits, TV and radio broadcasts
and informational literature are the tried and true techniques
in campaigning, whether for products, political candidates or Here, Word Libraries are used to assign the polarity. The
causes. Online campaigning had been gaining in popularity words are classified as positive, negative and neutral. We
and produced its large scale success with the elections of 2014. would use SentiWordNet in our system.
The score is assigned as +1 for positive, -1 for negative and
IV. S ENTIMENT A NALYZER : P ROPOSED S YSTEM 0 neutral. The overall score determines whether the tweet is
The proposed system works in 3 modules a positive response or a negative outrage. This score is stored
in the database for the corresponding tweet.
• Data Caching
• Text Pre-processing The scores for all the tweets related to the given topic can
• Sentiment Analysis be used for statistical study and predictions.
V. C ONCLUSION
In this paper we have seen various steps used to perform
sentiment analysis. We propose a system to analyze social
media contents in real time while collecting and efficiently
managing social media contents. We can use sentiment polarity
measures for various campaigns to see how positively or nega-
tively people react or talk about them. Analyzing the sentiment
of tweets gives an interesting insight into the opinions of the
public in relation to a certain events or campaigns. We can
see that results are encouraging as we are able to segregate
sentiments as positive, neutral and negative. Since the system
can analyze shared social media contents freely used in real
life in real time to find events related to campaigns or election
poll, and analyze and predict sentimental paths, the system
can be utilized for elections during the election period. It can
be used for diverse marketing programs.
VI. F UTURE S COPE
The system we are implementing can be used in various
fields like business, elections, for any kind of products and
services. Further the module will include to explore the
possibilities of accuracy improvement by introducing Machine
Learning Techniques for analysis and applying different vari-
ants of optimization methods. Moreover, there is a scope of
improvement for dealing with sarcasm and irony tweets. In
future, word level and post level contextual information along
with domain specific ontology can also be considered for the
classification of the tweets.
ACKNOWLEDGMENT
We would like to take the opportunity to express my sincere
thanks to our guides and B.E. teachers for providing advice
and valuable guidance.
We also extend our sincere thanks to all the faculty members
and the non-teaching staff and friends for their cooperation.
Last but not the least, we are thankful to all our family
members whose constant support and encouragement in every
aspect helped us to complete our project.
R EFERENCES
[1] Bouazizi, Mondher, and Tomoaki Ohtsuki. ”Opinion min-
ing in Twitter: How to make use of sarcasm to enhance
sentiment analysis. ” Advances in Social Networks Analysis
and Mining (ASONAM), 2015 IEEE/ACM International
Conference on.. IEEE, 2015.
[2] M. Trupthi, S. Pabboju, and G. Narasimha, Sentiment
analysis on twitter using streaming api,” in Advance Com-
puting Conference (IACC), 2017 IEEE 7th International),
pp. 915–919, IEEE, 2017.

You might also like