International Journal of Scientific Research in Engineering and Management (IJSREM)

Volume: 09 Issue: 01 | Jan - 2025 SJIF Rating: 8.448 ISSN: 2582-3930

Twitter sentiment Analysis Using Machine Learning

Ashwini S.Domade 1, Snehal K.Gadakh2, Vaishnavi G.Kardile 3 , Damini S.Pekhale 4

1
Ashwini Domade Department of Information and Technology from Matoshri Aasrabai Polytechnic
2
Snehal Gadakh Department of Information and Technology from Matoshri Aasrabai Polytechnic
3
Vaishnavi Kardile Department of Information and Technology from Matoshri Aasrabai Polytechnic
4
Damini Pekhale Department of Information and Technology from Matoshri Aasrabai Polytechnic
5
Ms.Vidya Kale lecturer of Information Technology from Matoshri Aasrabai Polytechnic
6
Mr.Mahesh Bhandakkar Head of Information Technology from Matoshri Aasrabai Polytechnic

---------------------------------------------------------------------***--------------------------------------------------------------------
Abstract - With the advancement of web technology more than 206 million daily active users in 2022, which is
defined as the number of logged accounts that can be iden-
and its growth, there is a huge volume of data present in tified by the platform and where ads can be shown. As more
the web for internet users and a lot of data is generated people contribute to social media, the analysis of informa-
tion available online can be used to reflect on the changes in
too. Internet has become a platform for online learning, people's perceptions, behavior, and psychology (Alamoodi et
exchanging ideas and sharing opinions. Social al. 2021). Hence, using Twitter data for sentiment analysis has
become a popular trend. The growing interest in social media
networking sites like Twitter, Facebook, Google+ are analysis has brought more attention to Natural Lan- guages
rapidly gaining popularity as they allow people to share Processing (NLP) and Artificial Intelligence (AI) technologies
related to text analysis.
and express their views about topics, have discussion
with different communities, or post messages across the Using text analysis, it is possible to determine the
senti- ments and attitudes of certain target groups. Much of
world. There has been lot of work in the field of the available literature focuses on texts in English but there is
sentiment analysis of twitter data. This survey focuses a growing interest in multilanguage analysis (Arun and
Srinagesh 2020a; Dashtipour et al. 2016; Lo et al. 2017). Text
mainly on sentiment analysis of twitter data which is analysis can be done by extracting subjective com- ments
helpful to analyze the information in the tweets where toward a certain topic using different sentiments such as
Positive, Negative, and Neutral (Arun and Srinagesh 2020b).
opinions are highly unstructured, heterogeneous and are One of the topical interests would be related to the
either positive or negative, or neutral in some cases. In Coronavirus (Covid-19), which is a novel disease that was
first discovered in late 2019. The rapid spread of Covid-19
this paper, we provide a survey and a comparative
worldwide has affected many countries, leading to changes in
analyses of existing techniques for opinion mining like people’s lifestyles, such as wearing masks on public
transportation and maintaining social distancing. Sentiment
machine learning and lexicon-based approaches, together
analysis can be implemented to social media data to explore
with evaluation metrics. Using various machine learning changes in people’s behavior, emotions, and opinions such as
by dividing the spread trend of Covid-19 into three stages and
algorithms like Naive Bayes, Max Entropy, and Support
exploring people’s negative sentiments toward Covid-19
Vector Machine, we provide research on twitter data based on topic modeling and feature extraction (Boon-Itt and
streams. We have also discussed general challenges and Skunkan 2020). Previous studies have retrieved tweets based
on certain hashtags (#) used to categorize content based on
applications of Sentiment Analysis on Twitter. certain topics such as “#stayathome” and “#socialdistanc- ing”
to measure their frequency (Saleh et al. 2021). Another study
Key Words: Twitter, Sentiment analysis (SA), has used the Word2Vec technique and machine learn- ing
models, such as Naive Bayes, SVC, and Decision Tree, to
Opinion mining, Machine learning, Naive Bayes (NB), explore the sentimental changes of students during the online
Maximum Entropy, Support Vector Machine (SVM). learning process as various learning activities were moved
online due to the pandemic (Mostafa 2021).

1.INTRODUCTION
Social media platform such as Twitter provides a
space where users share their thoughts and opinion as well as 2. PROBLEM STATEMENT
con- nect, communicate, and contribute to certain topics using With the explosive growth of social media,
short, 140 characters posts, known as tweets. This can be done
platforms like Twitter have become a vital source for
through texts, pictures, and videos, etc., and users can interact
using likes, comments, and reposts buttons. Accord- ing to capturing public sentiment on various topics ranging
Twitter (https://investor.twitterinc.com), the platform has from products and services to political events and social

© 2025, IJSREM | www.ijsrem.com DOI: | Page 1

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 09 Issue: 01 | Jan - 2025 SJIF Rating: 8.448 ISSN: 2582-3930

issues. However, the sheer volume and unstructured 3. Convolutional Neural Networks for Sentence
nature of this data make it challenging to extract Classification
meaningful insights. Companies, governments, and
organizations often struggle to manually analyze and This research explores the use of Convolutional Neural
understand public opinion in real-time, missing Networks (CNNs) for sentiment classification. It shows that
opportunities to respond to trends or public concerns. CNNs outperform traditional approaches in handling short
texts like tweets by effectively capturing semantic nuances.
This project aims to address the challenge of
automatically analyzing sentiments from Twitter data. 4. Sentiment Analysis and Opinion Mining
The system will classify tweets into positive, negative, or
neutral categories using machine learning techniques. Liu provides a broad review of sentiment analysis techniques,
The solution will help businesses, political organizations, especially their application to social media platforms. The
or individuals make data-driven decisions based on paper outlines the challenges of working with Twitter data and
public opinion by providing real-time sentiment analysis offers solutions like lexicon-based approaches and advanced
on Twitter data. Additionally, the system will visualize machine learning models.
sentiment trends, providing a clearer understanding of the
5. A System for Real-time Twitter
sentiment landscape around specific topics, events, or
brands. The authors present a real-time sentiment analysis system
3.LITERATURE REVIEW used to track public opinion during the 2012 U.S. Presidential
Election. The study demonstrates the effectiveness of real-
1] Twitter as a Corpus for Sentiment Analysis and time sentiment tracking but also discusses the importance of
Opinion Mining data quality and volume in maintaining accuracy.
Numerous sentiment analysis solutions utilize machine
learning and natural language processing techniques for 6. Deep Learning in Sentiment Analysis:
analyzing social media data, particularly Twitter. Studies have
Recent advancements in deep learning have further improved
demonstrated that systems like sentiment classifiers or opinion
sentiment classification performance. Studies like Socher et
mining tools are effective in classifying data into categories
al. (2013) used Recursive Neural Networks (RNN) and
(positive, negative, neutral). However, many systems struggle
Convolutional Neural Networks (CNN) to achieve state-of-
with real-time analysis, handling of noisy data (e.g., slang,
the-art results on sentiment analysis tasks. These models can
emojis), and managing large volumes of streaming data from
capture contextual and semantic information from text,
platforms like Twitter. Furthermore, existing solutions are
making them ideal for handling complex sentiment patterns in
often limited in their ability to detect nuanced sentiments such
tweets
as sarcasm and irony.

Challenges:
Research reveals several challenges in Twitter sentiment
analysis systems, including:

1) Noisy Data: Tweets often contain informal language,

special characters, and abbreviations.
2) Short Text Length: The 280-character limit on
tweets limits context and makes classification
difficult.
3) Sarcasm & Irony: Detecting sarcasm is difficult
with basic NLP techniques.
4) Real-Time Processing: Managing and analyzing
real-time data streams from Twitter requires robust
and scalable systems.
4.PROPOSED SYSTEM

2. Sentiment Analysis of Twitter Data

This paper develops a sentiment analysis model tailored to
Twitter, focusing on how unique features like emoticons and
hashtags contribute to sentiment. The study finds that
accounting for these features enhances classification accuracy .

© 2025, IJSREM | www.ijsrem.com DOI: | Page 2

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 09 Issue: 01 | Jan - 2025 SJIF Rating: 8.448 ISSN: 2582-3930

fore, all contractions in the tweets are expanded into their

formal forms, such as “isn’t” become “is not.”
(5) Clearing all non-alphabetical characters or symbols
including punctuation, numbers, and other special sym-
bols that may affect the feature extraction of the text.
(6) Removing duplicated or empty tweets and creating a
clean dataset.
(7) Converting emojis to their real meaning as many Twit-
ter users use emojis in their tweets to express their
sentiments and emotions. Hence, using the demojize()
function in the emoji module of Python and trans-
forming emojis into their true meaning may improve
the accuracy of the sentiment analysis (Tao and Fang
2020).

Fig: Distribution of collected tweets based on the selected 5. Application and Future Scope of the Project
cities and different stages
1.Applications that use Reviewsfrom Websites:
Today Internet has a large collection of reviews and
the data were pre-processed before sentiment analysis feedbacks on almost everything. This
(Naseem et al. 2021). We implemented a basic data-cleaning includes product reviews, feedbacks on political issues,
process as follows: comments about services, etc.

(1) Replacing upper-case letters to avoid recognizing the 2. Applications as a Sub-component Technology:-
same word as different words because of A sentiment predictor system can be helpful in recommender
capitalization. systems as well. The recommender system will not
(2) Removing hashtags (#topic), mentioned usernames recommend items that receive a lot of negative feedback or
(@username), and all the links that start with “www,” fewer ratings.
“http,” and “https.” Removing stop words and short
words (less than two characters). The stop words are 3. Applications in Business Intelligence:-
mostly very common in the text but hardly contain any It has been observed that people nowadays tend to look upon
sentiment polarity. However, in sentiment analy- sis, reviews of products which are available online before they
“not” and “no” should not be listed as stop words, buy them. And for many businesses, the online opinion
because removing these negations would change the decides the success or failure of their product.
real meaning of entire sentences.
(3) Reducing repeated characters from some words. Some 4. Applications across Domains:
users will type repeated characters to express their Recentresearches in sociology and other fields like medical,
strong emotions, so these words that are not in the lexi- sports have also been benefitted by Sentiment Analysis that
cons should be converted into their corresponding cor- show trends in human emotions especially on social media.
rect words. For example: “sooooo goooood” becomes
“so good.” 5. Applications In Smart Homes Smart homes
(4) Expanding contractions in tweets such as “isn't” or are supposed to be the technology of the future. In future
“don't” as these will become meaningless letters or entire homes would be networked and people would be able to
words after punctuations have been removed. There- control any part of the home using a tablet device.

Scopes :-

1]This project focuses on sentiment analysis for tweets in

English.

2]It will use supervised learning algorithms such as Logistic

Regression, Support Vector Machine (SVM), or Naive Bayes
for classification.

3]The system can be extended to analyze sentiment trends

over time and track changes in public opinion.

4]It can be useful for businesses, political organizations, or

individuals to understand public sentiment towards a product,
event, or policy.

© 2025, IJSREM | www.ijsrem.com DOI: | Page 3

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 09 Issue: 01 | Jan - 2025 SJIF Rating: 8.448 ISSN: 2582-3930

Comput Eng. https://doi. org/10.11591/ijece.v10i6.pp5992-

6. CONCLUSIONS 6000
7. Baccianella S, Esuli A, Sebastiani F (2010)
In conclusion, this paper extracts data regarding Covid-19 SENTIWORDNET 3.0: an enhanced lexical resource for
from people in the main cities of England on Twitter and sentiment analysis and opinion mining. In: Proceedings of
separates it into three different stages. First, we perform data the 7th international conference on language resources and
cleaning and use unsupervised lexicon-based approaches to evaluation, LREC 2010
classify the sentiment orientations of the tweets at each stage.
Then, we apply the supervised machine learning approaches 8. Barbosa L, Feng J (2010) Robust sentiment detection
using a sample of annotated data to train the Random Forest on twitter from biased and noisy data. In: Coling 2010—
23rd interna- tional conference on computational linguistics,
classifier, Multinomial Naïve Bayes classi- fier, and SVC, proceedings of the conference, 2
respectively. From lexicon-based approaches, the three stages 9. Bonta V, Kumaresh N, Janardhan N (2019) A
of public sentiment changes about the Covid-19 pandemic can comprehensive study on Lexicon based approaches for
be found. For most cities, the pro- portion of positive sentiment analysis. Asian J Comput Sci Technol 8(S2):1–6.
sentiments increases first and then drops, while the proportion https://doi.org/10.51983/ajcst- 2019.8.s2.2037
10. Boon-Itt S, Skunkan Y (2020) Public perception of the
of negative sentiments changed in a different direction. In COVID-19 pandemic on Twitter: sentiment analysis and
addition, by analyzing the number of deaths and confirmed topic modeling study. JMIR Public Health and Surveillance,
cases as well as vaccination situations, it could be concluded 6(4), e21978. https://doi.org/10.2196/21978
that the increase in confirmed cases and the decrease in 11. Breiman L (2001) Random forests. Mach Learn.
vaccination volume might be the reason for the increase in https://doi.org/10.
negative sentiments, even though further research is needed to 12. 1023/A:1010933404324
confirm this inference. For lexicon-based approaches, the
existing lexicon is

modified to better fit the language habits of modern social

media, improving the accuracy of this approach. Addition-
ally, an annotated dataset can be created to compare the dif-
ference between predicted results and real results. Research on
Covid-19 can be based on time series so that the changes in
people’s attitudes and perceptions can be analyzed over some
time. Moreover, further studies can combine the senti- ment
classification results with other factors such as deaths and
vaccination rates and establish a regression model to analyze
which factors contribute to the sentiment changes. Overall, the
paper has showcased different methods of con- ducting
sentiment analysis with SVC using BoW or TF–IDF
outperformed the model accuracy overall.

REFERENCES

1. Adwan OY, Al-Tawil M, Huneiti AM, Shahin RA, Abu

Zayed AA, Al-Dibsi RH (2020) Twitter sentiment analysis
approaches: a survey. Int J Emerg Technol Learn.
https://doi.org/10.3991/ijet. v15i15.14467
2. al Amrani Y, Lazaar M, el Kadirp KE (2018) Random
forest and support vector machine based hybrid approach
to sentiment analysis. Proc Comput Sci.
https://doi.org/10.1016/j.procs. 2018.01.150
3. Alamoodi AH, Zaidan BB, Zaidan AA, Albahri OS,
Mohammed KI, Malik RQ, Almahdi EM, Chyad MA,
Tareq Z, Albahri AS, Hameed H, Alaa M (2021) Sentiment
analysis and its applica- tions in fighting COVID-19 and
infectious diseases: a system- atic review. Expert Syst
Appl. https://doi.org/10.1016/j.eswa. 2020.114155
4. Aqlan AAQ, Manjula B, Lakshman Naik R (2019) A study
of senti- ment analysis: Concepts, techniques, and
challenges. In Lecture notes on data engineering and
communications technologies, vol 28.
https://doi.org/10.1007/978-981-13-6459-4_16
5. Arun K, Srinagesh A (2020a) Multi-lingual Twitter
sentiment analy- sis using machine learning. Int J Electr
Comput Eng. https://doi. org/10.11591/ijece.v10i6.pp5992-
6000
6. Arun K, Srinagesh A (2020b) Multi-lingual Twitter
sentiment analy- sis using machine learning. Int J Electr

© 2025, IJSREM | www.ijsrem.com DOI: | Page 4