A BRIEF REVIEW OF SENTIMENT ANALYSIS METHODS

International Journal of Information Sciences and Techniques (IJIST) Vol.6, No.1/2, March 2016
DOI : 10.5121/ijist.2016.6210 89
A BRIEF REVIEW OF SENTIMENT ANALYSIS
METHODS
Shreya Banker1
and Rupal Patel2
1
Smt. Chandaben Mohanbhai Patel Institute of Computer Applications, CHARUSAT,
CHANGA
2
Smt. Chandaben Mohanbhai Patel Institute of Computer Applications, CHARUSAT,
CHANGA
ABSTRACT
On today’s fast spreading use of social media, many websites have offer reviews of items like books, cars,
mobiles, movies etc. They describe the product in some detail and evaluate them as good/bad, preferred/not
preferred, so it is necessary to categorize these reviews in an automated way. Sentiment analysis is one
kind of computational technique of Artificial Intelligence. It is the task of identifying positive and negative
opinions, emotions, and evaluations. This article represents Sentiment analysis, it’s issues, applications and
some of the methods used to evaluate the review using sentiment analysis.
KEYWORDS
Sentiment analysis, Review Mining, Opinion Mining, Data mining, Machine learning
1. INTRODUCTION
Human opinion is important for making decision or choosing from multiple options. Before the
advent of the Internet, people were generally asking to friends or relatives for the product
recommendations. Nowadays, with the growing use of social media, people have access to refer
reviews on a different websites to take decision.
People’s opinion and experience are very valuable in decision-making process, but to get profits
from these opinion and experience, the accumulated content should be analysed properly. These
analysed opinions are useful for consumer to get a chance to evaluate others view and
involvement related to some products or services before purchasing them. Likewise, the trades-
person can take this review as feedback from the consumers, and they can improve the quality of
their product or services.
Sentiment analysis also called “Opinion Mining.” It is useful in several ways, such as social
media monitoring. There are also many names and slightly different tasks, e.g., sentiment analysis
[1], opinion mining [2], opinion extraction [3], sentiment mining [4], subjectivity analysis [5],
affect analysis [6], emotion analysis [7], review mining [8], etc. Sentiment analysis involves
classifying opinions into categories like "positive" or "negative” view. Approaches to this task
mainly drawn from the areas of natural language processing [9], data mining [10], and machine
learning [11].

90
2. LITERATURE REVIEW
There are different methods to perform sentiment analysis, such as keyword spotting, lexical
affinity and statistical methods.
Sentiment analysis includes a process called keyword spotting that includes developing a list of
keywords that relate to a certain sentiment. These words, known as affect words, are usually
positive or negative adjectives because such words can be strong indicators of sentiment.
Keyword spotting classifies text by affect categories based on the presence of unambiguous affect
words such as happy, sad, afraid, and bored. Keyword spotting method used to detect fraud.
Fraud examiners search for affect words in employee emails or other communications as part of
their fraud detection procedures. For example, a fraud examiner might receive a clue that a
manager is manipulating his division's reported sales to meet company quotas. In evaluating the
merit of the clue, the examiner might run a list of keywords — flexible, unreasonable, temporary
and worried — against the alleged party's emails and, if there are positive results investigate the
issues further. The limitation of this approach lies in two areas: poor recognition of affects when
negation is involved and reliance on surface features. About its first weakness, while the approach
can correctly classify the sentence “today was a happy day” as being happy, it is likely to fail on a
sentence like “today wasn’t a happy day at all.” About its second weakness, the approach relies
on the presence of obvious affect words that are only surface features of the prose.
Lexical affinity is slightly more sophisticated than keyword spotting as, rather than simply
detecting obvious affect words, it assigns arbitrary words a probabilistic ‘affinity’ for a particular
emotion. For example, ‘accident’ might assign a 75% probability of being indicating a negative
effect, as in ‘car accident’ or ‘hurt by accident’. Though often outperforming pure keyword
spotting, there are some problems with the approach. Lexical affinity, operating solely on the
word-level, can easily be tricked by sentences like “I avoided an accident” (negation) and “I met
my girlfriend by accident” (other word senses). [23].
Statistical methods, such as Bayesian inference and support vector machines, have been popular
for affect classification of texts. By feeding a machine learning algorithm a large training corpus
of affectively annotated texts, it is possible for the system to not only learn the affective valence
of affect keywords (as in the keyword spotting approach), but also to take into account the
valence of other arbitrary keywords (like lexical affinity), punctuation, and word co-occurrence
frequencies. The polarity of a word can be identified by studying the occurrence frequency of the
word in a large annotated corpus of texts [27]. If the word occurs more frequently among positive
texts, then its polarity is positive. If it occurs more frequently among negative texts, then its
polarity is negative. If it has equal frequencies, then it is a neutral word. The similar opinion
words frequently appear together in a corpus. This is the main observation that the state of the art
methods are based on. Therefore, if two words appear together frequently within the same
context, they are likely to have the same polarity. Therefore, the polarity of an unknown word can
be determined by calculating the relative frequency of co-occurrence with another word.
However, traditional statistical methods are generally semantically weak, meaning that, with the
exception, of obvious affect keywords, other lexical or co-occurrence elements in a statistical
model have little predictive value individually. As a result, statistical text classifiers only work
with acceptable accuracy when given a sufficiently large text input. Therefore, while these
methods may be able to affectively classify user’s text on the page- or paragraph- level, they do
not work well on smaller text units such as sentences or clauses [23].

91
3. PRELIMINARIES OF SENTIMENT ANALYSIS
Following are the preliminaries required for the sentiment analysis as:
3.1. Natural Language Processing
With the amount of information that is shared on social media, forums, blogs, etc., it is easy to see
why we need to automate sentiment analysis: there is simply too much information to manually
process. The problem is that machine-learning approaches are typically not that accurate. For a
simple task, separating ‘positive’ from ‘negative’ sentiment on social media, many automated
solutions only perform with around 80% accuracy. This can still be useful for tracking broad
trends over time, but it limits fine-grained analysis.
We need the natural language processing techniques to have a deeper understanding of the
context. Compare these examples:
1. It was a great restaurant.
2. It should have been a great restaurant.
3. The restaurant was great in that it will make all future meals seem more delicious.
4. Despite a pleasant experience, I cannot support the many reviews that it was a great
restaurant.
The first sentence is positive, but the rest are all negative sentiment, despite having the first
sentence embedded within them. The second sentence “should have been” indicates the desired
outcome, leaving the actual sentiment implied (it is possible that this sentence could follow by
“… and it was!”, but the implication is that it was not). The third is the more complicated,
bordering on the kinds of sarcasm that are very hard for machines to identify: even a person might
misread this as positive when skimming quickly. The fourth statement is even more complicated
because the overall sentiment is negative, but it begins with the (weakly) positive “pleasant
experience” and also finishes by reporting that many other people expressed positive sentiment.
3.2. Opinion Mining
We use the following review about a Canon camera to introduce the problems: Posted by: John
Smith Date: September 10, 2011
1. I bought a Canon G12 camera six months ago.
2. I simply love it.
3. The picture quality is amazing.
4. The Battery life is also long.
5. However, my wife thinks it is too heavy for her.
From this review, we identify some important issues as:

92
The review has a number of opinions, both positive and negative, about Canon G12 camera.
Sentence (2) expresses a positive opinion about the Canon camera as a whole. Sentence (3)
expresses a positive opinion about its picture equality. Sentence (4) expresses a positive opinion
about its battery life. Sentence (5) expresses a negative opinion about the weight of the camera.
From these opinions, we can make the following important observations:
An opinion consists of two key components: a target g and a sentiment s on the target. i.e. (g, s)
where g can be any entity or aspect of the entity about which an opinion has been expressed, and
s is a positive, negative, or neutral sentiment, or a numeric rating score expressing the
strength/intensity sentiment analysis and opinion mining 18 of the sentiment (e.g., 1 to 5 stars).
Positive, negative, and neutral are called sentiment (or opinion) orientations (or polarities). For
example, the target of the opinion in sentence (2) is Canon
G12, and the target of the opinion in sentence (3) is the picture quality of Canon G12. Target also
called topic in the literature.
This review has opinions from two persons, which are called opinion sources or opinion holders.
The holder of the opinions in sentences (2), (3), and (4) is the author of the review
(“John Smith”), but for sentence (5), it is the wife of the author. 3. The date of the review is
September 10, 2011. This date is important in practice because one often wants to know how
opinions change with time and opinion trends.
There are other factors to currently stop us from relying blindly on tools for sentiment analysis:
Context: A positive or negative sentiment word can have the opposite connotation depending on
context.
Sentiment Ambiguity: A sentence with a positive or negative word does not necessarily express
any sentiment (e.g., “can you recommend a good tool I could use?” does not express any
sentiment, although it uses the positive sentiment word “good“). Likewise, sentences without
sentiment words can express sentiment too (e.g., “This browser uses a lot of memory” does not
contain any sentiment words, although it’s clearly negative at a document level.).
Sarcasm: A positive or negative sentiment word can switch sentiment if there is sarcasm in the
sentence (e.g. “Sure, I’m happy for my browser to crash right in the middle of my coursework”).
Language: A word can change the sentiment and meaning depending on the language used.
This is often seen in slang, dialects, and language variations. An example is the word “sick,”
which can change meaning based on context, tone, and language, although clear to the target
audience.
4. LEVELS OF SENTIMENT ANALYSIS
Sentiment analysis has been handled as a Natural Language Processing task at many levels of
granularity. Depending on whether the target of study as a whole text or document, one or several
linked sentences, or one or several entities or aspects of those entities, different NLP, and
sentiment analysis tasks can be performed. Hence, it is necessary to distinguish three levels of
analysis that will clearly determine the different tasks of sentiment analysis:

93
Document Level: Analysing the overall sentiment expressed in the text. This works best on the
assumption that the whole comment only discusses one topic. Document level considers that a
document is an opinion on an entity or aspect of it. This level is associated with the task called
document-level sentiment classification. The task is to classify whether a whole opinion
document expresses a positive or negative sentiment. For example, given a product review, the
system determines whether the review expresses an overall positive or negative opinion about the
product. This task is known as document-level sentiment classification.
Sentence Level: The task at this level goes to the sentences and determines whether each
sentence expressed a positive, negative, or neutral opinion. Neutral usually means no opinion.
This level of analysis is closely related to subjectivity classification that distinguishes sentences
(called objective sentences) that express factual information from sentences (called subjective
sentences) that express subjective views and opinions. However, we should note that subjectivity
is not equivalent to sentiment as many objective sentences can imply opinions.
Aspect Level: Both the document level and the sentence level analysis do not discover what
exactly people liked and did not like. Aspect level performs finer-grained analysis. Aspect level
was called feature level (feature-based opinion mining and summarization).Aspect level directly
looks at the opinion itself. It is based on the idea that an opinion consists of a sentiment (positive
or negative) and a target (of opinion). It is closely related to tasks like Feature-based Opinion
Mining and Opinion Summarization.
5. APPLICATIONS OF SENTIMENT ANALYSIS
As the online review sites, blogs, social networking sites provide huge amount of opinions, this
helps in easier decision-making process. Some of the applications of sentiment analysis are
discussed as below:
Most of the e-commerce activities use sentiment analysis. Most of the websites allow users to
give their feedback about shopping and quality of products. They give the information about
different features and summary of product by giving ratings to products. Then, the users can view
reviews, read opinions about products and their special features. Moreover, summary of product
and its features is present to users in graphical form. Most of the popular commercial websites
like amazon.com give the information about reviews from customers with rating information. By
analysing these large volumes of opinions, Sentiment Analysis helps the websites by changing
dissatisfied customers into promoters [24].
Twitter and Facebook are a focal point of many sentiment analysis applications. The most
common application is monitoring the reputation of a specific brand on Twitter and/or Facebook.
One application that performs real-time analysis of tweets that contain a given term is tweetfeel
(http://www.tweetfeel.com) [25].
Voice of the Customer helps in determining what the individual customer is saying about
products or services. It means examining the reviews, opinions, and feedback of the customers.
Voice of the Customer is an important part of Customer Experience Management. To invent new
products, VOC helps in finding such new opportunities. Extracting customer opinions also helps
identify functional requirements of the products and some non-functional requirements like
performance and cost [24].

94
Sentiment analysis helps government in assessing their strength and weaknesses by analyzing
opinions from public. For example, “If there is situation like, The MP who is investigating 2G
scam himself is deeply corrupt then how do you expect that truth will 4 come out?.” This example
shows the negative sentiment about government. Sentiment analysis also helps in identifying
strengths and weaknesses in a recruitment campaign in government job, assessing success of
electronic submission of tax returns, and many other areas [24].
Sentiment analysis can provide substantial value to candidates running for various positions. It
enables campaign managers to track how voters feel about different issues and how they relate to
the speeches and actions of the candidates.[25]
Another important domain for sentiment analysis is the financial markets. There are numerous
news items, articles, blogs, and tweets about each public company. A sentiment analysis system
can use these various sources to find articles that discuss the companies and aggregate the
sentiment about them as a single score that can be used by an automated trading system. One such
system is The Stock Sonar.[25]
6. CONCLUSIONS
Sentiment analysis has a wide variety of applications in information systems, including
classifying reviews, summarizing review and other real time applications. Sentiment classifiers
are dependent on topics. It has found that different types of features and classification algorithms
combined in an efficient way in order to overcome their individual drawbacks, benefit, and finally
enhance the sentiment classification performance. In future, more work is needed on further
improving the performance. The techniques and algorithms used for sentiment analysis are
advancing fast; however, many problems in this field of study remain unsolved. More future
research could be dedicated to these challenges.
ACKNOWLEDGEMENTS
We would like to express our gratitude to the many people who saw us through this article; to all
those who provided support, allowed us to quote their remarks and assisted in the editing,
proofreading and design.
We would like to thank Dr. Atul Patel for enabling us to publish this paper and who encouraged
us lots throughout the entire survey.
We would like to thank Dr. Pritpal Singh for helping us in the process of selection and editing.
Special thanks to Mr. Kanu Patel our publisher who encouraged and guided us during the whole
process.
REFERENCES
[1] Wikipedia Retrieved from https://en.wikipedia.org/wiki/Sentiment_analysis
[2] Sentiment Mining Retrived from http://searchbusinessanalytics.techtarget.com/definition/opinion-
mining-sentiment-mining
[3] Retrived from https://www.cs.uic.edu/~liub/FBS/sentiment-analysis.html
[4] Retrived from https://www.cs.uic.edu/~liub/FBS/SentimentAnalysis-and-OpinionMining.pdf
[5] Retrived from http://www.aclweb.org/anthology/H05-2018.pdf
[6] Retrived from http://dl.acm.org/citation.cfm?id=1972851

95
[7] Retrived from https://www.kairos.com/emotion-analysis-api
[8] Retrived from https://www.cs.uic.edu/~liub/publications/kdd04-revSummary.pdf
[9] Retrived from https://en.wikipedia.org/wiki/Natural_language_processing
[10] Retrived from https://en.wikipedia.org/wiki/Data_mining
[11] Retrived from https://en.wikipedia.org/wiki/Machine_learning
[12] Li Zhuang, FengJing,Feng Jing. "Movie Review Mining and Summarization" (2006): 43-50.
[13] BakhtawarSeerat, FarouqueAzam. "Opinion Mining: Issues and Challenges (A survey)."
International Journal of Computer Applications (0975 – 8887) 49 (2012): 1-10.
[14] Vivek Narayanan, IshanArora, Arjun Bhatia. "Fast and accurate sentiment classification using
an enhanced Naive Bayes model" : 1-8.
[15] AnasCollomb, CrinaCostea,Damien Joyeux,Omar Hasan,Lionel Brunie. "A Study and
Comparison of Sentiment Analysis Methods for Reputation Evaluation."
[16] Bongirwar, Vrushali K. "A Survey on Sentence Level Sentiment Analysis." International Journal
of Computer Science Trends and Technology (IJCST) 3.3 (2015): 110-113.
[17] G.Vinodhini, RM.Chandrasekaran. "Sentiment Analysis and Opinion Mining: A Survey."
International Journal of Advanced Research in Computer Science and Software Engineering 2.6 (2002).
[18] Mr. Saifee Vohra, Prof. Jay Teraiya. "Applications and Challenges for Sentiment Analysis : A
Survey." International Journal of Engineering Research & Technology (IJERT) (2013): 1-5.
[19] Noura Farra, Elie Challita, Rawad Abou Assi, Hazem Hajj. "Sentence-level and Document-level
Sentiment Mining for Arabic Texts." 2010 IEEE International Conference on Data Mining Workshops
(n.d.).
[20] Pang, Bo. "AUTOMATIC ANALYSIS OF DOCUMENT SENTIMENT." (2006).
[21] V.K. Singh, R. Piryani, A. Uddin,P. Waila. "Sentiment Analysis of Movie Reviews." (n.d.): 712-
716.
[22] Wouter Bancken, Daniele Alfarone and Jesse Davis. "Automatically Detecting and Rating Product
Aspects from Textual Customer Reviews."1-16.
[23] Erik Cambria. "An Introductionto Concept-Level Sentiment Analysis."478-483.
[24] Sujata Rani "Rule Based Sentiment Analysis System".
[25] Ronen Feldman "TechniquesandApplicationsfor SentimentAnalysis" (2013).
[26] Aurangzeb Khan,Baharum Baharudin,and Khairullah Khan,Sentiment classification from online
customer reviews using lexical contextual sentence structure.pages 317-331.Springer,2011
[27] Read J, Carroll J. Weakly supervised techniques for domainindependent sentiment classification.
In: Proceeding of the 1st international CIKM workshop on topic-sentiment analysis for mass opinion; 2009.
p. 45–52.
Authors
Shreya Banker received her M.C.A. degree from G H Patel Post Graduate
Department of Computer Science and Technology, Sardar Patel University,
V.V.Nagar, Gujarat in 2010 & B.C.A. degree from SEMCOM College, Sardar Patel
University, V.V.Nagar, Gujarat in 2007. She is presently working as an Assistant
Professor in Smt. Chandaben Mohanbhai Patel Institute of Computer Applications,
Changa. Her research area includes Machine Learning and Data Mining
Rupal Patel received her M.C.A. degree from ISTAR, Gujarat Technological
University, V.V.Nagar, Gujarat, India in 2012, & B.C.A. degree from DDIT,
Dharmsinh Desai University, Nadiad, Gujrat, India in 2009. She is presently
working as an Assistant Professor in Smt. Chandaben Mohanbhai Patel Institute of
Computer Applications, Changa. Her research area includes Mobile Computing
and Database System.

A BRIEF REVIEW OF SENTIMENT ANALYSIS METHODS

More Related Content

Similar to A BRIEF REVIEW OF SENTIMENT ANALYSIS METHODS (20)

More from ijistjournal (20)

Recently uploaded (20)

A BRIEF REVIEW OF SENTIMENT ANALYSIS METHODS