Sentiment Analysis also known as opinion mining and Emotional AI
Refers to the use of natural language processing, text analysis, computational linguistics and biometrics to systematically identify, extract, quantify and study affective states and subjective information.
widely used in
Reviews
Survey responses
Online and social media
Health care
2. Sentiment Analysis also known as opinion mining and Emotional
AI
• Refers to the use of natural language processing, text analysis,
computational linguistics and biometrics to systematically
identify, extract, quantify and study affective states and subjective
information.
widely used in
• Reviews
• Survey responses
• Online and social media
• Health care
3. Oxford Dictionary Defines : Sentiment Analysis is process of
computationally identifying and categorising opinion expressed in piece
of text especially in order to determine whether the writer’s attitude
towards a particular topic, product etc is positive negative or neutral.
4. Characteristics of Sentiment Analysis
Sentiment
Categorization
Levels of
Analysis
Regular vs
comparative
opinion
Explicit vs
Implicit
Opinion
Role of
semantics
Dealing with
figures of
speech
Relationship in
Social
Networks
5. 1 Sentiment Categorisation
Objective Vs Subjective
The first aim when one is dealing with sentiment analysis usually consists in distinguishing
between subjective and objective sentences
If the sentence is classified as objective no other fundamental tasks are required while if
the sentence is subjective, its polarity (Positive ,negative or neutral) needs to be estimated.
Subjectivity classification is the tasks that distinguishes sentences that express objective or
factual information from the sentence that express subjective views and opinions.
Eg: objective “The iPhone is a smartphone
Subjective “The iPhone is awesome
7. 2 Levels of Analysis
Message Level
Sentence Level
Entity and Aspect
Level
• The aim is to classify polarity of a
whole opinioned message
• The aim is to determine the polarity
of each sentence contained in a text .
Single opinion on a single entity
• Performs a finer grade analysis than
others . It is on idea that an opinion
consists of a sentiment and target.
8. 3 Regular vs comparative opinion
Regular
Opinion
Indirect
Opinion
Direct Opinion
Regular opinion refers to literature as standard
opinion. It can be of 2 subtypes:
Direct opinion refers to opinion expressed
directly on an entity.
eg: The screen brightness of iPhone is awesome
Indirect opinion is an opinion that is expressed
indirectly on an entity on the basis of some other
entity
Eg: after I switched to iPhone I lost all my data.
9. Comparative Opinion
A comparative opinion expresses a relation of similarities or differences
between two or more entities or a preference of the opinion holder based
on some shared aspects of entities.
• Eg: ios is better performing than android and ios is best performing
operating system.
10. 4 Explicit vs Implicit Opinion
• Among the different shades that an opinion can assume.
Explicit and implicit:
Explicit Opinion: an explicit opinion is a subjective statement that implies a regular or
comparative opinion eg
Implicit Opinion : An implicit opinion is an objective statement that implies a regular or
comparative opinion tat usually expresses a desirable and undesirable fact.
11. 5 Role of Semantics
The semantics of language used in social networks in fundamental to
accurately analyse user expressions. The context of a textual
expression is therefore a crucial element should be taken into account
to properly deal with the underlying sentiment. A sentence “taken as it
is can appear as negative or positive ,but if it is properly analysed from
a semantic point of view it can be completely different.
12. 6 Dealing with Figures of Speech
• A Figure of speech is any artful deviation from the ordinary mode of
speaking or writing. In the tradition of Aristotle, figures of speech can
be divided into two groups- schemes and tropes. The function of
schemes and tropes. The function of schemes and tropes is to carry
out a transference of some kind, schemes are characterised by a
transference in meaning
13. 7 Relationships in Social Networks
Sentiment Analysis in Social Network is generally based on the assumption that
the text provided by the user are independent and identically distributed.
The first tentative approach is to deal with the real nature of social network
content is related to the principle of homophily eg: Friendship relationships can be used to
infer that connected users may be likelier to hold similar opinions, However a sentiment
analysis system should take into account that the assumptions about the friendships
relationships does not properly reflect the real world where two connected users have
difference of opinion
14. Types of Sentiment Analysis Method
• Document Level: This technique assumed that every document contains an opinion or
opinion related words. classification of these documents is carried out by using
approaches –supervised learning and Unsupervised learning.
• Sentence level: This technique is based on specialisation of documents which contains
various sentences and sentence contains an opinion behind it
• Aspect based : Attribute based classification is performed such as in customer reviews a
mobile phone which have different attributes like battery, camera etc. Each product
opinion is considered for better understanding sentiments
15. • Comparative sentiment Analysis : Identification of sentences which contain comparative
opinions and filter out and opinions are extracted by considering sequential patterns as features
• Sentiment Lexicon Acquisition: This is most popular and crucial resource for most of the
sentiment analysis algorithms. In this technique, acquisition of the sentiment lexicon is performed
by manually, dictionary-based approaches and corpus-based approaches
16. Importance of Sentiment Analytics in Social
Media Analytics
Determine Marketing Strategy
Improve campaign success
Improve product messaging
Improve customer Service
Test business KPI’s
Generate Leads
17. Types of sentimental Analysis
• Human interpretation of sentiment is definitely the most mature and
accurate judge of sentiment
Manual
Processing
• Keyword processing algorithms assign a degree of positivity or
negativity to an individual word. Eg: positive: love, great; negative:
dislike, Terrible
Keyword
processing
• NLP refers to computer systems that process human language in terms of
its meaning. Sentences conveys ideas. NLP works as a language for its
meaning. It converts speech to text ,language translation and grammar
checks
Natural Language
processing