SENTIMENTAL

ANALYSIS
➢ Sarthak Palande
➢ Pritviraj Shankar
➢ Ashish Kumar
➢ Girish Thorwade
Introduction
➢ Sentiment analysis is the process of using natural language processing,
text analysis, and statistics to analyze customer sentiment.
➢ It is the automated process of determining whether a text expresses a
positive, negative, or neutral opinion about a product or topic.
➢ The best businesses understand the sentiment of their customers—what
people are saying, how they’re saying it, and what they mean.
➢ Customer sentiment can be found in tweets, comments, reviews, or
other places where people mention your brand.
➢ Sentiment Analysis is the domain of understanding these emotions with
software, and it’s a must-understand for developers and business leaders
in a modern workplace.
Literature Survey
Abstract
➢ The world we see nowadays is becoming more digitalized. In this digitalized world E-commerce has
become more popular with the growth in internet and network technologies. Many people feel
convenient to buy products online using various forums such as Amazon, Flipchart, Awok etc.
➢ As now a day’s people are relying on online products so the importance of a review is going higher.
For selecting a product, a customer needs to go through thousands of reviews to understand a
product. Such product reviews are rich in information consisting of feedback shared by users.
➢ The review comments are useful to both other buyers and vendors. The sentiment analysis of
customer reviews helps the vendor to understand user’s perspectives. They can further use the
review comments and improve their When customers buy the products online there is an option for
them to provide their review comments.
➢ Many customers chose to provide their experience, opinion, feedback etc. In this project, the
correlation between the Amazon product reviews and the rating of the products given by the
customers is studied using Support Vector Machines (SVM) Machine Learning algorithm.
➢ By comparing these results, it could get a better understanding of the algorithm, and can also act as a
supplement to other fraud scoring detection methods.2.
Motivation
➢ Recent years have seen an increasing
amount of research efforts expanded
in understanding sentiment in textual
resources. The papers published on
sentiment analysis have been
increasing for the past years.
➢ One of the subtopics of this research is
called sentiment analysis or opinion
mining, which is, given a bunch of
text, it can computationally study
people’s opinions, appraisals,
attitudes, and emotions toward
entities, individuals, issues, events,
topics and their attributes.
Applications of this technique are
diverse.
➢ For example, businesses always want
to find public or consumer opinions
and emotions about their products and
services. Potential customers also
want to know the opinions and
emotions of existing users before they
use a service or purchase a product.
➢ Last but not least, researchers use this
information to do an in-depth analysis
of market trends and consumer
opinions, which could potentially lead
to a better prediction of the stock
market.
Purpose

➢ The objective of this paper is to classify the positive and negative reviews
of the customers over different products and build a supervised learning
model to polarize large amounts of reviews.
➢ This dataset consists of customers’ reviews and ratings, which was
extracted from Consumer Reviews of Amazon products. The extracted
features of this dataset and built several supervised model based on that.
➢ To compared the accuracy of these models and got a better understanding
of the polarized attitudes towards the products.
Scope
➢ The rise of social media has fueled
interest in sentiment classification.
Promptly and correctly classifying
sentiment from the text has become an
important task for individuals and
companies.
➢ In the development of prediction models
to classify the reviews, more reliable
approaches are expected to reduce the
misclassifications.
➢ In this study, the results of various hybrid
methods are empirically evaluated on
datasets of different size for use in
sentiment mining.
➢ To handle imbalance data distribution in
real time applications, it is observed from
the results that using SVMs for class
prediction can be influenced by the data
imbalance, although SVMs can adjust
itself to some degree of data imbalance.
➢ To cope with the problem, rebalancing
the data is chosen as a promising
direction, but both under sampling and
over sampling have limitations.
➢ Through extensive experiments with
benchmark and real application datasets,
the proposed modified bagging method is
shown to be effective and superior to
several other methods with different data
sampling methods.
Issues in current system
Implicit Sentiment and Sarcasm:-
➢ In sentiment analysis the system usually
predicts by training through the data
from a certain dataset.
➢ The issue with current dataset is that it
finds it difficult to predict if someone
has written a review sarcastically. For
example if someone writes “It is too
good to work smoothly.”
➢ In this context it can be seen that the
user clearly tells others that this device
won’t work but he used “too good to”
which by systems’ means is positive.
➢ This kind of reviews need very huge
amount of detailing so that the system
can recognize them.
Domain Dependency: -
➢ There are many words whose polarity
changes from domain to domain.
Consider the following examples. The
story was unpredictable.
➢ The steering of the car is unpredictable.
Go read the book. In the first example,
the sentiment conveyed is positive
whereas the sentiment conveyed in the
second is negative.
➢ The third example has a positive
sentiment in the book domain but a
negative sentiment in the movie domain
(where the director is being asked to go
and read the book).
Thwarted Expectations: -
➢ Sometimes the author deliberately sets
up context only to refute it at the end.
Consider the following example: This
film should be brilliant. It sounds like
a great plot, the actors are first grade,
and the supporting cast is good as
well, and Stallone is attempting to
deliver a good performance.
➢ However, it can’t hold up. In spite of
the presence of words that are positive
in orientation the overall sentiment is
negative because of the crucial last
sentence, whereas in traditional text
classification this would have been
classified as positive as term
frequency is more important there than
term presence.
Negation: -
➢ Handling negation is a challenging task in SA.
Negation can be expressed in subtle ways even
without the explicit use of any negative word.
➢ A method often followed in handling negation
explicitly in sentences like “I do not like the
movie”, is to reverse the polarity of all the
words appearing after the negation operator
(like not).
➢ But this does not work for “I do not like the
acting but I like the direction”. So to consider
the scope of negation as well, this extends only
till but here. So the thing that can be done is to
change polarity of all words appearing after a
negation word till another negation word
appears. But still there can be problems.
Benefits of Support Vector Machine (SVM) Method
➢ Support vector machines (SVM) are supervised learning method that can be used for solving
sentiment classification problems (Cristianini & ShaweTaylor 2000). This technique is based on
a decision plane where labeled training data is placed and then algorithm gives an optimal hyper
plane which splits the data into different groups or classes. As seen in above figure .The best
hyper plane is the one that separates the classes with the largest margin. This is achieved by
choosing a hyper plane so that its distance from the nearest data on each class is maximized
(Berk 2016).
➢ SVM’s are very good when there is no idea on the data. Works well with even unstructured and
semi structured data like text, Images and trees. The kernel trick is real strength of SVM. With
an appropriate kernel function, any complex problem solved. Unlike in neural networks, SVM is
not solved for local optima. It scales relatively well to high dimensional data. SVM models have
generalization in practice; the risk of over-fitting is less in SVM. SVM is always compared with
ANN. When compared to ANN models, SVMs give better results.
➢ Many software tools are available for SVM implementation. SVMs are really good for text
classification. SVMs are good at finding the best linear separator. The kernel trick makes SVMs
non-linear learning algorithms. Choosing an appropriate kernel is the key for good SVM and
choosing the right kernel function is not easy. Patient is key while building SVMs on large
datasets. They take a lot of time for training.
Need Of Machine Learning in
Sentimental Analysis
➢ The main objective of machine learning
is to enhance and increase the text
analytics capabilities that semantic
analysis does, also include the Role of
Speech tagging.
➢ Data scientists run a machine learning
model to recognize the actual sense by
providing a large volume of text
documents containing a lot of pre-tagged
description.
➢ When the model is ready, the same data
scientist can apply those training methods
to building new models to identify other
elements of speech.
➢ The result is very quick and that reliable
Part of Speech tagging helps in a bigger
text analytics system to identify
sentiment-bearing phrases more
effectively and accurately.
➢ Machine learning also helps in
information analysts to solve tricky
problems caused by the growth of
language.
➢ Creating some sentiment analysis rule set
for such type of platform where the thing
is impossible.
Objectives
➢ Collecting data i.e. product reviews on various websites featuring
various things.
➢ Analyze and categorize review data.
➢ Analyze sentiment on dataset from document level (review level).
➢ Categorization or classification of opinion sentiment into-
○ Positive
○ Negative
➢ Main Objective is to automate the process of determining whether a
text expresses a positive, negative, or neutral opinion about a product
or topic.
A Typical Sentimental Analysis Model
Conclusion
➢ Sentiment analysis deals with the
classification of texts based on the
sentiments they contain.
➢ This article focuses on a typical
sentiment analysis model consisting
of three core steps, namely data
preparation, review analysis and
sentiment classification, and
describes representative techniques
involved in those steps.
➢ Applications that utilize results
from sentiment analysis is also
expected to emerge in the near
future.
➢ Sentiment analysis is an emerging
research area in text mining and
computational linguistics, and has
attracted considerable research
attention in the past few years.
➢ Future research shall explore
sophisticated methods for opinion
and product feature extraction, as
well as new classification models
that can address the ordered labels
property in rating inference.
References
➢ https://readwrite.com/2020/06/02/understanding-the-role-of-ai-and-ml-in-sentiment-analysis/
➢ https://www.rcciit.org/
➢ https://www.researchgate.net/publication/236203597_Sentiment_Analysis_A_Literature_Survey
THANK YOU