Sentiment analysis and opinion mining is almost same thing however there is minor difference between them that is opinion mining extracts and analyze people's opinion about an entity while Sentiment analysis search for the sentiment words/expression in a text and then analyze it. It uses machine learning techniques like SVM (Support Vector Machines) to analyze the text and classify them as positive, negative or neutral.

1. OPINION MINING
- SENTIMENT ANALYSIS
PRESENTED BY:
RACHANA RAVEENDRAN
CSE IV YEAR
14H51A05A8
CMRCET

2. INTRODUCTION
• Today due to vast use internet and social platforms,
people are having a huge amount of space where they
can publically express their opinions.
• Some formal reviews are also available in various
discussion forums related to products/sites or domains.
• These opinions are directly related to how they feel. And
this feeling can be classified as being positive, negative
or neutral in nature.

3. WHAT IS OPINION MINING?
• Opinion mining is a combination of natural language
processing and text mining.
• The main objective of Opinion mining is Sentiment
Classification i.e. to classify the opinion into positive or
negative classes.
• It is a system involving collecting and examines the text
about any event from different sources like comments,
reviews, posts, tweets.

4. =
Basic components of an opinion are:
1. Opinion holder: The person or organization that
holds a specific opinion on a particular object.
2. Object: The item on which an opinion is expressed.
3. Opinion: A view, attitude, or appraisal on an object
from an opinion holder.

5. FOR EXAMPLE -
• For example, if the word ''small'' is used for the size of
electronic device then it is consider as the positive
statement, whereas if the statement consist "small"
word for the height of somebody then it is consider as
a negative statement.
• For example, the reviews of Americans may vary from
the Indian about the President of USA or a movie
reviews may vary from location to location.

6. TYPES OF TEXTS
• In opinion mining, evaluation of text is of two types,
Direct and Comparison.
• In direct opinion the statement is direct and the
sentiments are independent from other object.
• For example ‘x’ phone has good features.
• Whereas in comparison opinion, object in one
statement compared with the other object.
• For example ‘x’ phone is better than ‘y’ phone.

7. CLASSIFICATION
• Classification of opinions is done on the basic two
levels
 Sentence level
 Document level
• In sentence level classification, we can classify the
sentences as objective or subjective.
• In document level classification, we can classify the
sentences as more opinionated.

8. SENTENCE LEVEL
• First task is to determine whether the statement is
subjective or objective and then classify the sentence
and determine if it is positive, negative or neutral.
• Riloff and Wiebe proposed a method named as
bootstrap method.
• In Bootstrap, sentences are labeled by two classifier,
high confidence objective sentences and high
confidence subjective sentences.
• Rest are left unlabeled.
• Subjective classifier looks for the list whereas objective
classifier locates sentences without those words.
• After this phase, patterns are formed of same category
and analyzed by machine learning approaches.

9. DOCUMENT LEVEL
• The objective is to determine whether the document is
positive or negative about a certain object.
• On the basis of polarity, the pre-selected word is
analyzed and the approach is based on the distance
measure of adjectives found in text.
• The adjectives are extracted that provides contextual
information.
• Then, the semantic orientation is measured.
The semantic orientation of an opinion on a feature states
whether the opinion is positive, negative or neutral.

10. METHOD USED
• Support vector machine (SVM) is the supervised
learning approach used for the classification.
• It is the most significant method for the classification
in supervised learning and better than the Naive Bayes
for the text categorization.
• It provide large margin between classifiers and the
basic idea to use SVM is to find the Hyper plane which
is not only separating two classes but also gives the
maximum margin between them.

11. • Select the hyper-plane which
segregates the two classes
better.
• The margin for hyper-plane C is high as compared to both A and B.
Hence, we name the right hyper-plane as C.

12. Select the hyper-plane that has maximum margin. Hence, we can say, SVM is
robust to outliers.

13. APPLICATIONS AND
BENEFITS
• Applications of SA range from public voice analysis,
crowd surveillance, customer care to exploit the
publics’ online content generation for analyzing inputs
such as emotion and responses towards local events.
• It can be used to give your business valuable insights
into how people feel about your product brand or
service.
• When applied to social media channels, it can be used
to identify spikes in sentiment, thereby allowing you to
identify potential product advocates or social media
influencers.

14. APPLICATIONS AND
BENEFITS
• It can be used to identify when potential negative
threads are emerging online regarding your business,
thereby allowing you to be proactive in dealing with it
more quickly.
• Sentiment analysis could also be applied to your
corporate network, for example, by applying it to your
email server, emails could be monitored for their
general “tone”.
• For example, Tone Detector is an Outlook Add-in that
determines the “tone” of your email as you type. Like
an emotional spell checker for all of your outgoing
email.

15. IBM WATSON’S
EMOTION DETECTION
Hi Team,
The times are difficult! Our sales have been disappointing
for the past three quarters for our data analytics product
suite. We have a competitive data analytics product suite
in the industry. However, we are not doing a good job at
selling it, and this is really frustrating.
We are missing critical sales opportunities. We cannot
blame the economy for our lack of execution. Our clients
are hungry for analytical tools to improve their business
outcomes. In fact, it is in times such as this, our clients
want to get the insights they need to turn their
businesses around. It is disheartening to see this
happening. Let’s buckle up and execute.

16. After processing the email via IBM Watson’s Tone Analyzer,
here are the results:

17. DIFFICULTIES
• Sentiment analysis can be applied to many areas but
arriving at whether a statement is positive or negative
can be difficult.
• The human language can be complex for machine
based learning systems to interpret.
• For example, opinions can be expressed with sarcasm
or irony, and the order of words can add even more
confusion.
“I currently use the Nikon D90 and love it, but not as much
as the Canon 40D/50D. I chose the D90 for the video
feature. My mistake.”

18. DIFFICULTIES
• Sometimes the value of the retrieved data is not
realized immediately and therefore the issue of how
long to store the data requires the attention of both,
the data centre officers as well as the strategic
planning units.
• Generally the data keeps growing and hence data
management issues such as it’s storage structure,
accessibility control, warehousing and compressing will
rise.
• SA techniques should also be updated to be able to
reason and determine the levels of uncertainty, validity,
messiness and trustworthiness of the data

19. FUTURE WORK
• Studies in opinion mining approaches have existed for
more than a decade and now are exploited by
enterprises as an important tool for strategic marketing
planning and manoeuvring.
• It is predicted that studies and skills development on
opinion mining for brand monitoring and customer
relation management are going to get increasing
attention and thus promising of higher revenues for
companies.
• Furthermore, brand management approaches through
opinion mining are expanding and creating a
marketing tsunami in many organizations, thus shifting
focus towards personalization and a consumer-centric
engagement.

20. CONCLUSION
• Sentiment analysis has a wide variety of application in
summarizing reviews, classifying reviews, information
system, market analysis and decision making.
• Sentiment analysis is a broad range of fields of natural
language processing and text mining.
• It is found that different types of features and
classification techniques are combined in an efficient
way to enhance the sentiment classification.
• It is also able to collect useful information from the
social sites, blogs and micro blogging site thus
benefitting various organizations.

21. REFERENCES
• https://www.analyticsvidhya.com/blog/20
17/09/understaing-support-vector-
machine-example-code/
• http://thescipub.com/PDF/jcssp.2016.153
.168.pdf
• https://socialmediaweek.org/blog/2017/0
8/4-machine-learning-emotion-detection-
apis-need-try/
• Opinion Mining and Analysis: A
Literature Review by Vandana Singh and
Sanjay Kumar Dubey (IEEE Paper)

22. THANK YOU!