A Survey on Analysis of Twitter Opinion Mining using Sentiment Analysis

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
© 2017, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 1307
A Survey on Analysis of Twitter Opinion Mining Using Sentiment
Analysis
Anusha K S1 , Radhika A D2
1M Tech, CSE Dept. of Computer Science and Engineering VVCE, Mysuru
2Assistant Professor, CSE Dept. of Computer Science and Engineering VVCE, Mysuru
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - In recent years, there is a rapid growth in online
communication. There are many social networking sites and
related mobile applications, and some more are still
emerging. Huge amount of data is generated by these sites
everyday and this data can be used as a source for various
analysis purposes. Twitter is one of the most popular
networking sites with millions of users. There are users with
different views and varieties of reviews in the form of tweets
are generated by them. Nowadays Opinion Mining has
become an emerging topic of research due to lot of
opinionated data available on Blogs & social networking
sites. Tracking different types of opinions & summarizing
them can provide valuable insight to different types of
opinions to users who use Social networking sites to get
reviews about any product, service or any topic. Analysis of
opinions & its classification on the basis of polarity (positive,
negative, neutral) is a challenging task. Lot of work has been
done on sentiment analysis of twitterdata andlotneedsto be
done.
In this paper we discuss the levels, approaches of
sentiment analysis, sentiment analysis of twitter data,
existing tools available for sentiment analysis and the steps
involved for same. Two approaches are discussed with an
example which works on machinelearningandlexiconbased
respectively.
Keywords—Twitter data, opinion mining, sentiment
analysis.
I. INTRODUCTION
Sentimentanalysistechniqueisaneffectivemeansof
discovering public opinions. Various companies often use
online or paper based surveys tocollectcustomercomments.
Due to the emergence of social networking sites and
applications, people tend to comment on their facebook or
tweet profile. Therefore the paper based approach is not an
efficient approach. Only a very small customer base can be
reached and there is no guarantee that their answers in the
survey are honest or not. Here social media comes into play.
Facebook, Twitter and all other social media sites are full of
people’s opinions about products/services they use,
comments about popular personalities and much more [1].
Nasukawa & YI first introduced the term Sentiment Analysis
& Opinion Mining in the year 2003. Opinion mining toolswill
help in providing the opinion about the product. Sentiment
analysis on tweet data involves data collection, extraction,
classification, understanding and providing the opinion that
are expressed in various tweets [2].
Opinion Mining and Sentiment analysis is done on
three levels [3]
 Document Level: Analysis is done on the whole
document and then expresses whether the
document is positive or negative sentiment.
 Sentence Level: It is related to find sentiment
polarity from short sentences. Sentence level is
merely close to subjectivity classification.
 Entity /Aspect Level: sentiment analysis performs
augmented analysis. The aim is to find sentiment on
entities or aspects.
Two approaches of Sentiment Analysis
 Supervised approaches or machine learning
method:
Machine learning is one of the most prominent
techniques gaining researchers interest [10] due to
its adaptability and accuracy. This method
comprises of three stages:(i)Data collection(ii)Pre-
processing and(iii)Trainingdata Classification
[9].
 Unsupervised (or lexicon-based):
Lexical analysis estimates the sentiment from the
semantic orientation of words [8] or phrases that
occur in a text. In this approach a dictionary
containing positive and negative words that are
matched with the words containing in tweet.
However, these techniques totally depend onlexical
resources [6] which are concerned with mapping
words [7] to a categorical (positive, negative,
neutral) or numerical sentiment score. In this
method the unigrams, whicharefoundinthelexicon
[9] are assigned a polarity score.
Sentiment Analysis of twitter data
Twitter is popular online social networking service
launched in March 2006.It enables users to send and read
tweets with about 140 characters length. Currently Twitter
acts as opinionated Data Bank with large amount of data
available used for sentiment analysis. Twitter is very
convenient for research because there are very large
numbers of messages, many of which are publicly available,
and obtaining them is technically simple compared to
scarping blogs from the web.
Twitter data is collected for analysis using Twitter API.
Volume: 04 Issue: 12 | Dec-2017 www.irjet.net p-ISSN: 2395-0072

II.Tools Available for sentiment analysis
S. No Tools Applications
1 NLTK
NLTK toolkit is widely used nowadays
for sentiment analysis task. Main
features of NLTK used in Sentiment
analysis process are Tokenization, Stop
Word removal, Stemming and tagging.
This tool is written in Python language
and can be downloaded from
www.nltk.org.
2 GATE
General Architecture for Text
Engineering (GATE) is information
Extraction System, Stemming and Partof
speech tagger. This tool is written in Java
language. https://gate.ac.uk/
3
Red
Opal
This tool is widely used for users who
want to buy any products based on
different features. Users can search for
any product depending upon the feature
selected and can get reviews related to
their search.
4
Opinion
finder
Opinion Finder is used for analysis of
different Subjective sentences related to
any topic & classification of sentences is
done based on their polarity. It’s written
in Java and is platform Independent tool.
Table-1: Sentiment Analysis Tools
III.TWITTER SENTIMENT ANALYSIS PROCEDURE
The framework used for this analysis is depicted in
below figure. Different processing steps had their own
important role. We discussed about all steps below.
Figure-1 Sentiment analysis Framework
A. Data Collection:
Collection of data is an important part of Sentiment Analysis.
Various data Sources like Blogs, Review Sites, Online Posts &
Micro Blogging like Twitter, Facebook are used for Data
Collection.
B. Data Preprocessing:
Now before Sentiment Analysis we need to process the
collected data using the following steps of data processing-
1) Stemming- In this process thepostfixfromeachwords
like “ing”,“tion” etc are removed.
2) Tokenization- This process is very important for Data
pre processing as it includes several sub steps like
“Removal of Extra spaces”, “Emoticons (,) used
replaced with their actual meaning like Happy, Sad by using
Emoticon data set available on Internet”, “Abbreviations like
OMG, WTF are replaced by their actual meanings”,
“Pragmatics handling like hapyyyyyyy as happy, guddddd as
good etc.”
3) Stop Word Removal- In this, stop words which are not
of any use in analysis like Prepositions (a, an) and
Conjunctions (and, between) used are removed.
C. Feature Extraction:
Feature extraction specifies the type of features used for
opinion Mining [6]. There are different typesoffeaturesused
like-
1)Term Frequency- Frequency of any termina document
carries weight age. [6]
2)Term Co-occurrence- Repeatedly occurrence ofa word
like Unigram, Bigram or n-gram etc. [6]
3)Part of Speech-For each tweet we have features for
counts of the number of Verbs, adjectives, nouns. [7]
D. Sentiment Analysis & Polarity Classification:
Emotions,opinionsandsentimentsplayanimportantrole
in all human life. Mining such opinions termed as sentiment
analysis [10]. Performing task of Sentiment analysis &
polarity classification is a challengingtask.SentiWordnet isa
standard dictionary used by most researchers today for
sentiment analysis. Task of Polarity classification mean the
reviews collected areclassifieddependingupontheemotions
expressed as Positive, Negative and Neutral.

IV. APPROACHES USED FOR SENTIMENT
ANALYSIS ON TWITTER
A. SENTIMENT ANALYSIS ON TWITTER USING
SUPERVISED APPROACH (MACHINE LEARNING)
This approach extracts the data from SNS services which is
done using Streaming APIoftwitter.The extractedtweets are
loaded into hadoop and it is been preprocessed using map
reduce.This task is followed by classification whichusesNLP
and machine learning techniques. The classification used
here is uni-word naïve bayes’ classification.
FIGURE 2: real time sentiment analysis on twitter
Consider the number of all positive tweets, positive words
and negative words from our training phase. Then calculate
the probability of a tweet being positive.
P(C) = No. of Positive Tweets / Total Number of Tweets - (1)
For each word in each tweet that is being streamed is
checked for the probability of it being used given that it is
positive.
P(D/C) = Positive score of the words / Total number of
Positive Words - (2)
Then checked the word itself being used irrespective of
whether or not it is positive.
P(D) = Positive score of the word + Negative score of the
Word / Total Number of words - (3)
In-order to check the probabilityofwordbeing positivegiven
that is used in a tweet which is given as follows:
P(C/D) = P(C) * P(D/C) / P(D) - (4)
The probability of a word is then passed to the Sentiment
function which then classifies,iftheprobabilityofthe wordis
greater than 0.6then it is positive, asneutral iftheprobability
is between 0.4 and 0.6 and negative if it is lesser than 0.
B. SENTIMENT ANALYSIS ON TWITTER USING
UNSUPERVISED APPROACH (LEXICON METHOD)
The data is collected from the twitter API and that data is
pre-processed to eliminate all unwanted information and to
replace the emoticons. Here lexical method is used for
classification and work on dictionary-based approach. The
dictionary-based approach depends on finding words from
tweets, and then matches the word with the dictionary. If
there is a positive match, the positive score is shown or the
word is tagged as positive. If it is negative word then the
negative score is incremented or the word is tagged as
negative. Otherwise tag neutral word.
FIGURE 3: working of a lexical technique
Let a sentence s contains a set of entities {e1, e2, …, er} and a
subset of their aspects {a1, …, am} from a set of opinion
holders[2] {h1, h2, …, hp} at some particular time pointanda
set of sentiment words or phrases {sw1, …, swn} with their
sentiment scores . The sentiment orientation for each aspect
ai in s is determined by the following aggregation function:

Where dist(swj, ai) is the distance between aspect ai and
sentiment word swj in s. swj.so is the sentiment score of swi.
If the final score is positive, then the opinion on aspect ai in s
is positive. If the final score is negative,thenthesentiment on
the aspect is negative. It is neutral otherwise.
CONCLUSION
In this paper, we have done a short survey on sentiment
analysis and opinion mining. We have discussed about three
major levels of sentiment analysis, two approaches of
sentiment analysis and sentiment analysis of twitter data.
Further, we studied and listed some of the tools available for
sentiment analysis and the general procedure for sentiment
analysis. By analyzing the examples of both approaches we
understood the two approaches in detail.
REFERENCES
[1] Syed Akib Anwar Hridoy, M.Tahmid Ekram,
Mohammad Samiul Islam, Faysal Ahmed and
Rashedur M. Rahman “Localized twitter opinion
minimg using sentiment analysis”.
[2] Roshan Fornandes, Dr. Rio D’Souza “Analysis of
product twitter data through opnionmining”©2016
IEEE.
[3] M. Trupthi, Suresh Pabboju, G. Narasimha
“Sentiment analysis on twitter using streamingAPI”
2017 IEEE &th International Advance Computing
Conference.
[4] Prerna Mishra, Dr. Ranjana Rajnish, Dr. Pankaj
Kumar “Sentiment analysis of twitter data: Case
study on digital india” 2016(InCITe).
[5] Paramita Ray, Amlan Chakrabarti “Twitter
sentiment analysis for product reviews using
Lexicon Method” 2017(ICDMAI).
[6] A Kowcika and Aditi Guptha “sentiment Analysis for
social media” ,International journal of advanced
research in computer science and software
engineering,216-221,Volume 3,Issue 7, july 2013.
[7] G. Vinodini and RM.Chandrashekaran, “sentiment
analysis and opinion mining: A survey”,
International journal of advanced research in
computer science and software enginnering,283-
294, Volume 2,Issue 6, june 2012.
[8] Cataldo Musto, Giovanni Semeraro, Marco
Polignano, “A comparison of Lexicon-based
approaches for Sentiment Analysis of microblog
posts”, Department of Computer Science, University
of Bari Aldo Moro, Italy.
[9] James Spencer and Gulden Uchyigit, Sentimentor:
Sentiment Analysis of Twitter Data. School of
Computing, Engineering and Mathematics.
University of Brighton.
[10] Anna Jurek, Maurice D. Mulvenna and Yaxin Bi,
Improved lexiconbased sentimentanalysisforsocial
media analytics Science direct, Published: 9
December 2015.
[11] Apoorv Agarwal Boyi Xie Ilia Vovsha OwenRambow
Rebecca Passonneau, “SentimentAnalysisofTwitter
Data”, Columbia University, Newyork.
[12] Sang-Hyun Cho and Hang-Bong Kang, “Text
Sentiment Classification for SNS-based Marketing
Using Domain Sentiment Dictionary”, IEEE
International Conference on Conference on
consumer Electronics(ICCE), p.717-718, 2012.
[13] Patricia L V Ribeiro, Li Weigang and Tiancheng Li “A
Unified Approach for Domain-Specific Tweet
Sentiment Analysis”, FUSION, 2015.
[14] Tiara, Mira Kania Sabariah, Veronikha Effendy,
“Sentiment Analysis on Twitter Using the
Combination of Lexicon-Based and Support Vector
Machine for Assessing the Performance of a
Television Program”, 3rd International Conference
on Information and Communication Technology
(ICoICT), 2015.
[15] Asmita Dhokrat, Sunil Khillare, C. Namrata
Mahender, “Review on Techniques and Tools used
for Opinion Mining,” IJCAT, 2015.

A Survey on Analysis of Twitter Opinion Mining using Sentiment Analysis

More Related Content

What's hot (20)

Similar to A Survey on Analysis of Twitter Opinion Mining using Sentiment Analysis (20)

More from IRJET Journal (20)

Recently uploaded (20)

A Survey on Analysis of Twitter Opinion Mining using Sentiment Analysis