The American Journal of Applied Sciences
6
https://www.theamericanjournals.com/index.php/tajas
TYPE
Original Research
PAGE NO.
6-16
10.37547/tajas/Volume07Issue01-02
OPEN ACCESS
SUBMITED
16 October 2024
ACCEPTED
09 December 2024
PUBLISHED
07 January 2025
VOLUME
Vol.07 Issue01 2025
CITATION
Pinky Akter, Safayet Hossain, Md Tarake Siddique, Mohammad Iftekhar
Ayub, Ayan Nath, Paresh Chandra Nath, Mohammad Rasel, & Md
Mehedi Hassan. (2025). Sentiment Analysis of Consumer Feedback and
Its Impact on Business Strategies by Machine Learning. The American
Journal of Applied Sciences, 7(01), 6
–
16.
https://doi.org/10.37547/tajas/Volume07Issue01-02
COPYRIGHT
© 2025 Original content from this work may be used under the terms
of the creative commons attributes 4.0 License.
Sentiment Analysis of
Consumer Feedback and
Its Impact on Business
Strategies by Machine
Learning
Pinky Akter
1
, Safayet Hossain
2
, Md Tarake
Siddique
3
, Mohammad Iftekhar Ayub
4
, Ayan Nath
5
,
Paresh Chandra Nath
6
, Mohammad Rasel
7
, Md
Mehedi Hassan
8
1
Master Of Science in Information Technology, Washington University of
Science and Technology, USA
2
Master of Science in Cybersecurity, Washington University of Science and
Technology, USA
3
Master of Science in Information Technology, Washington University of
Science and Technology, USA
4
Master of Science in Information Technology, Washington University of
Science and Technology, USA
5
Master’s in computer and information science, International American
University, USA
6
Master of Science in Information Technology, Washington University of
Science and Technology, USA
7
Masters in Business Analytics, International American University, LA,
California, USA
8
Master of Science in Information Technology, Washington University of
Science and Technology, USA
Corresponding Author: Md Mehedi Hassan
Abstract:
Sentiment analysis is a powerful tool for
transforming consumer feedback into actionable
insights, enabling businesses to refine strategies and
improve customer experiences. This study evaluates the
performance of machine learning models, including
Logistic Regression, Random Forest, SVM, LSTM, and
BERT, for sentiment classification on a diverse dataset
of customer reviews. BERT outperformed other models,
achieving an AUC-ROC of 0.97 and an accuracy of 94.2%,
showcasing its ability to capture complex semantic
patterns in text. The findings provide businesses with
critical insights into consumer sentiment, guiding
decision-making and enhancing competitive advantage.
The study also addresses challenges such as data
ambiguity, ethical considerations, and computational
The American Journal of Applied Sciences
7
https://www.theamericanjournals.com/index.php/tajas
The American Journal of Applied Sciences
demands, offering practical recommendations for
implementing scalable and effective sentiment
analysis solutions. These results demonstrate the
potential of machine learning-driven sentiment
analysis in shaping customer-focused business
strategies and fostering growth in a data-driven
market.
Introduction:
Understanding consumer sentiment has
become an essential aspect of modern business
strategy. In an era where consumers engage
extensively through digital platforms, feedback
analysis provides businesses with actionable insights
into
customer
preferences,
complaints,
and
expectations. This process, known as sentiment
analysis, leverages natural language processing (NLP)
and machine learning algorithms to classify text data
into positive, neutral, or negative sentiments. The
ability to analyze consumer sentiment effectively
allows organizations to shape business strategies that
enhance customer experience, foster brand loyalty,
and optimize marketing campaigns.
Sentiment analysis plays a pivotal role in industries
such as e-commerce, hospitality, and finance, where
customer feedback directly impacts decision-making
and profitability. For example, customer reviews on
platforms like Amazon or Yelp often determine
product sales and brand reputation. Beyond customer
experience management, sentiment analysis also
influences broader business areas such as market
research, competitive intelligence, and public
relations. With the advent of advanced machine
learning models such as LSTM (Long Short-Term
Memory)
networks
and
transformer-based
architectures like BERT (Bidirectional Encoder
Representations from Transformers), the accuracy and
scalability of sentiment analysis have significantly
improved (Devlin et al., 2019).
Despite its growing importance, challenges remain in
the field, including the handling of ambiguous
language, domain-specific terminology, and cultural
nuances. Additionally, ensuring data quality and
developing robust validation mechanisms are critical
for reliable sentiment analysis. This study aims to
investigate the sentiment distribution in consumer
feedback and evaluate the performance of various
machine learning models, providing insights into their
practical applications in real-world business scenarios.
Literature Review
The field of sentiment analysis has witnessed
significant advancements in the past two decades,
primarily driven by improvements in computational
linguistics and artificial intelligence. Early methods
relied on rule-based systems and lexicons such as
SentiWordNet, which mapped words to predefined
sentiment scores (Esuli & Sebastiani, 2006). While
effective for simple tasks, these methods struggled with
the complexity of human language, including context,
sarcasm, and idiomatic expressions.
The introduction of machine learning brought a
paradigm shift, enabling the development of algorithms
that could learn from labeled datasets. Logistic
regression and support vector machines (SVM) became
popular techniques for text classification tasks, offering
improved accuracy overrule-based systems (Pang et al.,
2002). However, their reliance on manual feature
extraction limited their scalability and adaptability to
new domains.
Deep learning models marked the next milestone in
sentiment analysis. Recurrent Neural Networks (RNNs)
and their variants, such as LSTM, demonstrated superior
performance by capturing sequential dependencies in
text data (Hochreiter & Schmidhuber, 1997). These
models were particularly effective for long texts, where
context plays a crucial role in determining sentiment.
More recently, transformer-based models like BERT
have set new benchmarks in sentiment analysis. By
leveraging bidirectional context and pre-training on
massive text corpora, BERT achieves state-of-the-art
results in various NLP tasks, including sentiment
classification (Devlin et al., 2019).
The business applications of sentiment analysis are vast
and varied. In the retail sector, sentiment analysis helps
brands understand customer preferences and tailor
their offerings accordingly (Kotler & Keller, 2016). In the
financial industry, analyzing sentiments from news
articles and social media posts aids in market prediction
and risk assessment (Bollen et al., 2011). Furthermore,
sentiment analysis is increasingly being integrated into
customer relationship management (CRM) systems to
provide real-time insights and enhance customer
satisfaction (Sharma et al., 2020).
Despite these advancements, challenges persist.
Ambiguity in natural language, such as sarcasm or mixed
sentiments, remains difficult for even the most
advanced models to decipher. Additionally, the quality
of labeled datasets significantly impacts model
performance, emphasizing the need for robust data
validation techniques. Ethical considerations, such as
bias in training data and privacy concerns, also warrant
attention as sentiment analysis becomes more
pervasive.
This study builds upon existing research by comparing
the performance of traditional and modern machine
The American Journal of Applied Sciences
8
https://www.theamericanjournals.com/index.php/tajas
The American Journal of Applied Sciences
learning models for sentiment analysis. By analyzing
consumer feedback, we aim to highlight the practical
implications of sentiment analysis for business strategy
and contribute to the ongoing discourse on the role of
artificial intelligence in business intelligence.
Methodology
This study adopts a comprehensive methodology to
analyze consumer feedback through sentiment
analysis and evaluate its impact on business strategies.
The methodology is divided into several key
subsections,
including
data
collection,
data
preprocessing, sentiment analysis, data validation, and
business strategy evaluation. Each stage of the process
is described in detail to ensure clarity and
reproducibility of the research.
Data Collection
We collected a diverse dataset from three primary
sources: online customer reviews, social media
platforms, and customer surveys. These sources were
chosen for their ability to capture consumer
sentiments across various industries and touchpoints.
The data from e-commerce platforms such as Amazon
and eBay provided insights into consumer opinions
about products and services. Social media data from
platforms like Twitter and Facebook allowed us to
understand spontaneous and real-time customer
sentiments. Customer survey responses offered
structured feedback on customer satisfaction and
preferences.
The dataset consists of 50,00 feedback samples,
ensuring a balanced representation of positive, neutral,
and negative sentiments. To maintain the integrity and
diversity of the data, we included feedback from
multiple industries such as retail, healthcare, and
hospitality. Each feedback entry is labeled with key
attributes, including a unique identifier, the textual
content of the feedback, sentiment category, sentiment
score, source of the feedback, and the timestamp of
submission.
The data collection process involved the following steps:
1.
Web Scraping: We used Python-based libraries
such as BeautifulSoup and Scrapy to extract textual
feedback from e-commerce and social media platforms.
APIs like Twitter API were employed to fetch tweets
containing keywords related to customer experiences
and sentiments.
2.
Survey Integration: Customer surveys were
either obtained through collaborations with businesses
or created for this study. These surveys were distributed
online using platforms like Google Forms and
SurveyMonkey, targeting a diverse audience across
different industries.
3.
Data Cleaning: Duplicate entries, irrelevant
feedback, and spam content were removed to ensure
the dataset's quality. Feedback with insufficient textual
content or ambiguous sentiments was excluded.
4.
Ethical Compliance: All data was collected in
compliance with privacy policies and regulations. Social
media data was publicly available, and survey
participants provided informed consent for their
feedback to be used in this study.
Table 1: Dataset and Attributes
ATTRIBUTE
DESCRIPTION
TYPE
FEEDBACK_ID
A unique identifier assigned to each feedback entry
Numeric
FEEDBACK_TEXT
The textual content of consumer feedback, including reviews, posts, and
survey comments
Text
SENTIMENT_LABEL
The overall sentiment of the feedback categorized as Positive, Neutral, or
Negative
Categorical
SENTIMENT_SCORE
A numerical score ranging from -1 to +1, representing the intensity of
sentiment (e.g., -1 for strongly negative, +1 for strongly positive)
Numeric
SOURCE
The origin of the feedback, such as Amazon, Twitter, or Survey
Categorical
INDUSTRY
The industry to which the feedback pertains (e.g., Retail, Healthcare,
Technology)
Categorical
REGION
The geographic location of the feedback (e.g., USA, Europe, Asia)
Categorical
TIMESTAMP
The date and time when the feedback was submitted, allowing for
temporal trend analysis
Date/Time
CUSTOMER_ID
A unique identifier for the customer providing the feedback (anonymized
for privacy)
Numeric
PRODUCT_CATEGORY
The category of the product or service associated with the feedback
(e.g., Electronics, Apparel)
Categorical
The American Journal of Applied Sciences
9
https://www.theamericanjournals.com/index.php/tajas
The American Journal of Applied Sciences
This table illustrates the structure and key attributes
of the dataset, ensuring transparency and
organization in our approach.
Data Preprocessing
To prepare the raw textual data for sentiment
analysis, we followed a detailed preprocessing
protocol to enhance the quality and consistency of
the inputs. First, we cleaned the data by removing
punctuation, special characters, and unnecessary
whitespace. This step ensured that the feedback
text was free of irrelevant noise. We then split the
text into individual words or tokens using
tokenization, enabling more granular analysis.
To further refine the data, we removed common
stopwords, such as "is," "the," and "and," which do
not contribute significantly to the sentiment of the
feedback. We also applied stemming and
lemmatization to reduce words to their root forms,
thereby minimizing redundancy and improving the
consistency of textual representations. For
instance, words like "running," "ran," and "runs"
were reduced to their base form "run."
Additionally, categorical variables, such as
sentiment labels, were encoded into numerical
formats to facilitate machine learning analysis. The
preprocessing process ensured that the textual data
was transformed into a structured format suitable
for input into sentiment analysis models.
Sentiment Analysis
We conducted sentiment analysis by implementing
both machine learning and deep learning
techniques, ensuring that the feedback was
analyzed with precision and depth. The primary goal
of this analysis was to classify consumer feedback
into three sentiment categories: positive, neutral,
and negative, and to assign a sentiment score to
quantify the strength of these sentiments.
The sentiment analysis began by utilizing traditional
machine learning models such as Support Vector
Machines (SVM). SVM was selected due to its
effectiveness in text classification tasks, particularly
in handling high-dimensional data. Additionally, we
employed Random Forests, a robust ensemble
learning method that combines multiple decision
trees to improve classification accuracy and reduce
the risk of overfitting.
To achieve more nuanced and sophisticated
analysis, we leveraged advanced deep learning
models, including BERT (Bidirectional Encoder
Representations from Transformers). BERT, a pre-
trained
transformer-based
model,
has
demonstrated state-of-the-art performance in
natural language processing tasks. We fine-tuned
BERT on our dataset to extract contextual
information from feedback text, allowing for a more
accurate understanding of sentiment nuances.
Each model was trained and tested on the
preprocessed dataset, with 70% of the data
allocated for training and 30% for testing.
Performance evaluation metrics, such as accuracy,
precision, recall, and F1-score, were calculated for
each model to compare their effectiveness in
classifying
sentiments.
Furthermore,
we
implemented cross-validation techniques to ensure
that the models generalized well to unseen data.
To enhance interpretability, we also generated
sentiment distribution visualizations, highlighting
the proportion of positive, neutral, and negative
feedback across different industries and feedback
sources. The insights gained from this sentiment
analysis formed the foundation for linking
consumer opinions to business strategies.
Data Validation
To ensure the reliability and integrity of the dataset
used in this study, we implemented rigorous data
validation processes. We began by verifying the
completeness of the dataset, ensuring that no
critical fields, such as feedback text or sentiment
labels, were missing. Any incomplete or irrelevant
entries were removed to maintain the quality of the
data.
We conducted duplicate checks to eliminate
redundant feedback entries, particularly in data
collected from social media platforms, where
reposts or repeated comments are common.
Additionally, we performed consistency checks to
ensure that sentiment labels aligned with the
textual content of the feedback. Mislabelled entries
were identified and corrected through manual
inspection or automated techniques using rule-
based algorithms.
Outlier detection was also performed to identify
and address any anomalies in the dataset, such as
extremely high or low sentiment scores that could
skew the analysis. Statistical methods and
visualization tools, such as box plots, were used to
identify these outliers.
To validate the accuracy of the sentiment labels, we
conducted inter-rater reliability tests, where
The American Journal of Applied Sciences
10
https://www.theamericanjournals.com/index.php/tajas
The American Journal of Applied Sciences
multiple reviewers independently labeled a subset
of the feedback. The agreement among reviewers
was measured using Cohen’s Kappa coefficient,
ensuring that the sentiment labels were consistent
and reliable. This step further strengthened the
credibility of the dataset and the subsequent
analysis.
Business Strategy Evaluation
The impact of consumer sentiments on business
strategies was assessed by mapping the results of
sentiment analysis to key business performance
indicators. We examined metrics such as customer
retention rates, sales growth, and brand reputation
to understand the strategic implications of
consumer feedback.
Using statistical methods such as regression
analysis, we identified correlations between
consumer sentiments and business performance
metrics. For example, we analyzed how negative
sentiments might correlate with a decline in
customer retention or how positive sentiments
could drive sales growth. We also applied time
series modeling to predict trends and gain insights
into how changing sentiment patterns could
influence future business decisions.
The insights gained from this analysis were used to
refine and optimize business strategies in critical
areas such as product development, marketing
campaigns, and customer service. By linking
sentiment trends to actionable business outcomes,
we ensured that the analysis was both practical and
impactful.
We ensured that all stages of the research adhered
to ethical guidelines. The data collected for this
study was either publicly available or obtained with
proper permissions from survey respondents. We
took necessary precautions to anonymize sensitive
information, ensuring that the privacy of individuals
was protected. For survey data, we sought explicit
consent from participants and adhered to
applicable data protection regulations. Additionally,
ethical approval was obtained where required to
ensure that our research methods met the highest
standards of integrity and transparency.
The research was conducted using Python
programming and its associated libraries. We used
Scikit-learn for machine learning tasks, TensorFlow
for implementing deep learning models, and Pandas
for efficient data manipulation. To visualize the
results and present insights effectively, we
employed Matplotlib and Seaborn. The use of these
tools ensured that our analysis was performed with
precision, scalability, and efficiency.
This methodological framework provides a robust
and comprehensive approach to analyzing
consumer feedback, extracting valuable insights,
and linking them to actionable business strategies.
By leveraging advanced technologies and adhering
to ethical principles, we aim to contribute to the
growing field of sentiment analysis and its practical
applications in business.
Results
This section presents the findings of our sentiment
analysis and its implications for business strategies.
The results were derived from applying machine
learning models to the dataset, analyzing customer
sentiment trends, and evaluating the relationship
between sentiment and business performance.
Model Performance
The sentiment analysis models were evaluated
using standard metrics such as accuracy, precision,
recall, F1-score, and AUC-ROC. Among the tested
models, the BERT-based transformer model
outperformed others in capturing nuanced
sentiment expressions in consumer feedback. The
following table summarizes the performance of
each model:
MODEL
ACCURACY (%)
PRECISION (%)
RECALL (%)
F1-SCORE (%)
AUC-ROC
LOGISTIC REGRESSION
85.3
83.5
82.9
83.2
0.88
RANDOM FOREST
88.7
87.9
86.5
87.2
0.91
SUPPORT VECTOR MACHINE
89.5
88.8
87.2
88.0
0.92
LSTM NEURAL NETWORK
91.4
90.5
90.1
90.3
0.94
BERT (TRANSFORMER)
94.2
93.6
93.4
93.5
0.97
The BERT model achieved the highest accuracy (94.2%) and the best balance of precision and recall, making it
the most reliable choice for analyzing complex sentiment patterns.
The American Journal of Applied Sciences
11
https://www.theamericanjournals.com/index.php/tajas
The American Journal of Applied Sciences
Sentiment Distribution
The overall distribution of sentiments in the dataset is summarized below:
SENTIMENT
COUNT
PERCENTAGE (%)
POSITIVE
25,000
50.0
NEUTRAL
15,000
30.0
NEGATIVE
10,000
20.0
The bar chart below visualizes the sentiment distribution across the dataset:
Chart 1: Sentiment Distribution
The bar chart illustrating the sentiment distribution.
Positive sentiments account for 50%, Neutral
sentiments for 30%, and Negative sentiments for
20% of the dataset.
Temporal Trends in Sentiment
Analysis of sentiment over the five-year period
revealed interesting trends:
•
Positive Sentiments: There was a steady
increase in positive sentiments, particularly
in the retail and technology sectors, which
reflects
improvements
in
customer
satisfaction and service quality.
•
Neutral Sentiments: Neutral feedback
remained relatively consistent over the
years, highlighting areas where businesses
failed to evoke strong positive or negative
emotions.
•
Negative Sentiments: Negative sentiments
showed a slight decline, particularly in the
healthcare
and
hospitality
sectors,
suggesting that businesses are addressing
common customer concerns.
Impact on Business Strategies
The sentiment analysis results provided actionable
insights for business strategies:
•
Positive
Feedback:
Positive
reviews
highlighted factors such as product quality,
customer
service,
and
affordability.
Businesses can leverage this feedback to
reinforce their strengths and market their
successes.
•
Negative Feedback: Common themes in
negative
feedback
included
delayed
delivery, unresponsive customer support,
and product defects. Addressing these
issues promptly can help businesses
improve customer retention.
•
Neutral Feedback: Neutral reviews often
included suggestions for improvement or
general observations. Businesses should
treat this feedback as opportunities for
innovation and improvement.
Overall Results
The analysis demonstrates the significance of
sentiment analysis in understanding customer
needs
and
optimizing
business
strategies.
Businesses that actively monitor and respond to
The American Journal of Applied Sciences
12
https://www.theamericanjournals.com/index.php/tajas
The American Journal of Applied Sciences
customer feedback are more likely to achieve long-
term success and customer loyalty.
Her
e’s an overall bar chart summarizing the
performance of sentiment analysis models:
Chart 2 : Model Performance
The results underscore the importance of using
advanced machine learning models, such as BERT,
to analyze consumer feedback effectively.
Sentiment analysis provides critical insights that
empower businesses to enhance customer
experiences, resolve pain points, and maintain a
competitive edge. This study demonstrates the
potential of sentiment analysis as a valuable tool for
data-driven decision-making in business strategies.
The AUC-ROC curve showcasing the performance of
different models. The AUC-ROC values increase
progressively from Logistic Regression (0.88) to
BERT (0.97), illustrating the superior ability of
advanced models like LSTM and BERT to distinguish
between classes in sentiment analysis.
Explanation of the Curve
•
AUC-ROC (Area Under the Receiver
Operating Characteristic Curve) measures a
model's ability to differentiate between
positive and negative classes.
•
A higher AUC-ROC score indicates better
performance. For instance, an AUC-ROC of
0.97 (achieved by BERT) means the model is
The American Journal of Applied Sciences
13
https://www.theamericanjournals.com/index.php/tajas
The American Journal of Applied Sciences
97% effective in distinguishing between
classes.
•
Logistic Regression, while simpler, scores
0.88,
indicating
relatively
lower
effectiveness but still acceptable for basic
tasks.
Real-World Relevance
In real-world scenarios, the AUC-ROC metric is
crucial for understanding model performance,
particularly in applications where false positives and
false negatives have different consequences. For
example:
•
In healthcare, high AUC-ROC models ensure
accurate disease detection, minimizing
misdiagnoses.
•
In business, such as sentiment analysis, high
AUC-ROC models help accurately interpret
consumer sentiment, enabling better
decision-making and customer satisfaction.
•
In fraud detection, models with high AUC-
ROC values can reliably identify fraudulent
transactions while reducing false alarms.
DISCUSSION
The findings of this study highlight the critical role
of sentiment analysis in understanding consumer
feedback and shaping business strategies. By
leveraging advanced machine learning models such
as BERT, we demonstrated that modern NLP
techniques can achieve high levels of accuracy and
reliability in sentiment classification. The superior
performance of BERT, with an AUC-ROC score of
0.97 and accuracy of 94.2%, underscores the
transformative potential of deep learning in
sentiment analysis tasks. This result validates the
growing adoption of transformer-based models in
real-world business applications, where nuanced
understanding of customer sentiment is essential
for decision-making.
A notable observation is the incremental
improvement in performance across models, from
traditional approaches like logistic regression to
advanced neural architectures. While logistic
regression and Random Forest provided a solid
baseline, they were outperformed by SVM, LSTM,
and ultimately BERT. This performance progression
reflects the advancements in machine learning
techniques, particularly the ability of deep learning
models to capture contextual relationships and
semantic meaning in text. The consistent
improvement in accuracy and AUC-ROC values
demonstrates the importance of selecting
appropriate models for sentiment analysis,
depending on the complexity of the dataset and the
desired level of granularity.
From a practical perspective, the insights gained
from sentiment analysis can be applied across
various
business
domains.
For
instance,
understanding positive sentiments can help
companies identify strengths in their products or
services, while addressing negative sentiments can
guide improvements and damage control efforts. In
e-commerce, analyzing customer reviews can
optimize product recommendations and inventory
management. In financial services, sentiment
analysis of news and social media can provide early
indicators of market trends and investment
opportunities. These applications highlight the
strategic value of sentiment analysis in enhancing
customer satisfaction, increasing profitability, and
maintaining competitive advantage.
Despite these promising results, several challenges
persist. Data quality remains a significant concern,
as noisy or imbalanced datasets can negatively
impact model performance. Additionally, handling
ambiguous language, cultural nuances, and sarcasm
continues to be a limitation for even the most
advanced models. Ethical considerations, such as
mitigating bias in training data and ensuring
customer privacy, are equally important as
sentiment analysis becomes more prevalent in
business operations. Future research should
explore ways to address these challenges, including
the development of domain-specific models and
robust validation frameworks.
Another key observation is the scalability and
adaptability of the models. While BERT achieved the
highest performance metrics, it also demands
significant computational resources, which may not
be feasible for all organizations. Balancing model
performance with resource efficiency is an essential
consideration, particularly for small and medium-
sized enterprises. Lightweight models or transfer
learning
techniques
could
offer
practical
alternatives, allowing businesses to achieve
satisfactory results without incurring excessive
costs.
CONCLUSION
This study underscores the growing importance of
sentiment analysis as a tool for extracting
actionable insights from consumer feedback. By
evaluating the performance of various machine
learning models, we demonstrated that advanced
techniques like BERT outperform traditional
approaches, offering higher accuracy and reliability
The American Journal of Applied Sciences
14
https://www.theamericanjournals.com/index.php/tajas
The American Journal of Applied Sciences
in sentiment classification. These findings have
significant implications for businesses aiming to
leverage sentiment analysis to enhance customer
experience, optimize marketing efforts, and inform
strategic decision-making.
The study also highlights the importance of
selecting appropriate models and datasets to
achieve desired outcomes. While advanced models
provide superior performance, their computational
requirements may pose challenges for widespread
adoption. Therefore, businesses must weigh the
trade-offs between performance and feasibility
when implementing sentiment analysis solutions.
Furthermore, addressing challenges related to data
quality, ambiguity, and ethical considerations will
be critical for the continued advancement and
adoption of sentiment analysis technologies.
In conclusion, sentiment analysis offers immense
potential for transforming customer feedback into
valuable insights that drive business success. By
continuing to innovate and refine machine learning
techniques, businesses can unlock the full potential
of sentiment analysis, ensuring they remain
competitive in an increasingly data-driven world.
Future research should focus on overcoming
existing challenges, exploring new applications, and
ensuring ethical and responsible use of sentiment
analysis in real-world scenarios.
Acknowledgement: All the Authors contributed
equally
REFERENCE
Md Habibur Rahman, Ashim Chandra Das, Md
Shujan Shak, Md Kafil Uddin, Md Imdadul Alam,
Nafis Anjum, Md Nad Vi Al Bony, & Murshida Alam.
(2024). TRANSFORMING CUSTOMER RETENTION IN
FINTECH
INDUSTRY
THROUGH
PREDICTIVE
ANALYTICS AND MACHINE LEARNING. The
American Journal of Engineering and Technology,
6(10),
150
–
163.
https://doi.org/10.37547/tajet/Volume06Issue10-
17
Bollen, J., Mao, H., & Zeng, X. (2011). Twitter mood
predicts the stock market. Journal of Computational
Science,
2(1),
1-8.
https://doi.org/10.1016/j.jocs.2010.12.007
Devlin, J., Chang, M. W., Lee, K., & Toutanova, K.
(2019). BERT: Pre-training of deep bidirectional
transformers
for
language
understanding.
Proceedings
of
NAACL-HLT
2019.
https://doi.org/10.48550/arXiv.1810.04805
Esuli, A., & Sebastiani, F. (2006). SentiWordNet: A
publicly available lexical resource for opinion
mining. Proceedings of LREC 2006.
Hochreiter, S., & Schmidhuber, J. (1997). Long short-
term memory. Neural Computation, 9(8), 1735
–
1780. https://doi.org/10.1162/neco.1997.9.8.1735
Kotler, P., & Keller, K. L. (2016). Marketing
management (15th ed.). Pearson Education.
Pang, B., Lee, L., & Vaithyanathan, S. (2002).
Thumbs up?: Sentiment classification using machine
learning techniques. Proceedings of EMNLP 2002.
Sharma, A., Kumar, A., & Bhardwaj, R. (2020). Role
of sentiment analysis in improving customer
satisfaction. International Journal of Advanced
Research in Computer Science, 11(1), 12-18.
Tauhedur Rahman, Md Kafil Uddin, Biswanath
Bhattacharjee, Md Siam Taluckder, Sanjida Nowshin
Mou, Pinky Akter, Md Shakhaowat Hossain, Md
Rashel Miah, & Md Mohibur Rahman. (2024).
BLOCKCHAIN
APPLICATIONS
IN
BUSINESS
OPERATIONS AND SUPPLY CHAIN MANAGEMENT
BY MACHINE LEARNING. International Journal of
Computer Science & Information System, 9(11), 17
–
30.
https://doi.org/10.55640/ijcsis/Volume09Issue11-
03
Md Jamil Ahmmed, Md Mohibur Rahman, Ashim
Chandra Das, Pritom Das, Tamanna Pervin, Sadia
Afrin, Sanjida Akter Tisha, Md Mehedi Hassan, &
Nabila Rahman. (2024). COMPARATIVE ANALYSIS
OF MACHINE LEARNING ALGORITHMS FOR
BANKING FRAUD DETECTION: A STUDY ON
PERFORMANCE, PRECISION, AND REAL-TIME
APPLICATION. International Journal of Computer
Science & Information System, 9(11), 31
–
44.
https://doi.org/10.55640/ijcsis/Volume09Issue11-
04
Bhandari, A., Cherukuri, A. K., & Kamalov, F. (2023).
Machine learning and blockchain integration for
security applications. In Big Data Analytics and
Intelligent Systems for Cyber Threat Intelligence
(pp. 129-173). River Publishers.
Diro, A., Chilamkurti, N., Nguyen, V. D., & Heyne, W.
(2021). A comprehensive study of anomaly
detection schemes in IoT networks using machine
learning algorithms. Sensors, 21(24), 8320.
Nafis Anjum, Md Nad Vi Al Bony, Murshida Alam,
Mehedi Hasan, Salma Akter, Zannatun Ferdus, Md
Sayem Ul Haque, Radha Das, & Sadia Sultana.
(2024). COMPARATIVE ANALYSIS OF SENTIMENT
ANALYSIS MODELS ON BANKING INVESTMENT
IMPACT BY MACHINE LEARNING ALGORITHM.
International Journal of Computer Science &
Information
System,
9(11),
5
–
16.
The American Journal of Applied Sciences
15
https://www.theamericanjournals.com/index.php/tajas
The American Journal of Applied Sciences
https://doi.org/10.55640/ijcsis/Volume09Issue11-
02
Shahbazi, Z., & Byun, Y. C. (2021). Integration of
blockchain, IoT and machine learning for multistage
quality control and enhancing security in smart
manufacturing. Sensors, 21(4), 1467.
Das, A. C., Mozumder, M. S. A., Hasan, M. A.,
Bhuiyan, M., Islam, M. R., Hossain, M. N., ... & Alam,
M. I. (2024). MACHINE LEARNING APPROACHES FOR
DEMAND
FORECASTING:
THE
IMPACT
OF
CUSTOMER
SATISFACTION
ON
PREDICTION
ACCURACY. The American Journal of Engineering
and Technology, 6(10), 42-53.
Akter, S., Mahmud, F., Rahman, T., Ahmmed, M. J.,
Uddin, M. K., Alam, M. I., ... & Jui, A. H. (2024). A
COMPREHENSIVE STUDY OF MACHINE LEARNING
APPROACHES
FOR
CUSTOMER
SENTIMENT
ANALYSIS IN BANKING SECTOR. The American
Journal of Engineering and Technology, 6(10), 100-
111.
Shahid, R., Mozumder, M. A. S., Sweet, M. M. R.,
Hasan, M., Alam, M., Rahman, M. A., ... & Islam, M.
R. (2024). Predicting Customer Loyalty in the Airline
Industry: A Machine Learning Approach Integrating
Sentiment
Analysis
and
User
Experience.
International
Journal
on
Computational
Engineering, 1(2), 50-54.
Ontor, M. R. H., Iqbal, A., Ahmed, E., & Rahman, A.
LEVERAGING DIGITAL TRANSFORMATION AND
SOCIAL MEDIA ANALYTICS FOR OPTIMIZING US
FASHION BRANDS'PERFORMANCE: A MACHINE
LEARNING APPROACH. SYSTEM (eISSN: 2536-7919
pISSN: 2536-7900), 9(11), 45-56.
Rahman, A., Iqbal, A., Ahmed, E., & Ontor, M. R. H.
(2024). PRIVACY-PRESERVING MACHINE LEARNING:
TECHNIQUES,
CHALLENGES,
AND
FUTURE
DIRECTIONS IN SAFEGUARDING PERSONAL DATA
MANAGEMENT. International journal of business
and management sciences, 4(12), 18-32.
Md Jamil Ahmmed, Md Mohibur Rahman, Ashim
Chandra Das, Pritom Das, Tamanna Pervin, Sadia
Afrin, Sanjida Akter Tisha, Md Mehedi Hassan, &
Nabila Rahman. (2024). COMPARATIVE ANALYSIS
OF MACHINE LEARNING ALGORITHMS FOR
BANKING FRAUD DETECTION: A STUDY ON
PERFORMANCE, PRECISION, AND REAL-TIME
APPLICATION. International Journal of Computer
Science & Information System, 9(11), 31
–
44.
https://doi.org/10.55640/ijcsis/Volume09Issue11-
04
Arif, M., Ahmed, M. P., Al Mamun, A., Uddin, M. K.,
Mahmud, F., Rahman, T., ... & Helal, M. (2024).
DYNAMIC PRICING IN FINANCIAL TECHNOLOGY:
EVALUATING MACHINE LEARNING SOLUTIONS FOR
MARKET
ADAPTABILITY.
International
Interdisciplinary Business Economics Advancement
Journal, 5(10), 13-27.
Uddin, M. K., Akter, S., Das, P., Anjum, N., Akter, S.,
Alam, M., ... & Pervin, T. (2024). MACHINE
LEARNING-BASED EARLY DETECTION OF KIDNEY
DISEASE: A COMPARATIVE STUDY OF PREDICTION
MODELS AND PERFORMANCE
EVALUATION.
International Journal of Medical Science and Public
Health Research, 5(12), 58-75.
Das, A. C., Rishad, S. S. I., Akter, P., Tisha, S. A., Afrin,
S., Shakil, F., ... & Rahman, M. M. (2024).
ENHANCING
BLOCKCHAIN
SECURITY
WITH
MACHINE LEARNING: A COMPREHENSIVE STUDY OF
ALGORITHMS AND APPLICATIONS. The American
Journal of Engineering and Technology, 6(12), 150-
162.
Shak, M. S., Uddin, A., Rahman, M. H., Anjum, N., Al
Bony, M. N. V., Alam, M., ... & Pervin, T. (2024).
INNOVATIVE MACHINE LEARNING APPROACHES TO
FOSTER FINANCIAL INCLUSION IN MICROFINANCE.
International Interdisciplinary Business Economics
Advancement Journal, 5(11), 6-20.
Naznin, R., Sarkar, M. A. I., Asaduzzaman, M., Akter,
S., Mou, S. N., Miah, M. R., ... & Sajal, A. (2024).
ENHANCING SMALL BUSINESS MANAGEMENT
THROUGH MACHINE LEARNING: A COMPARATIVE
STUDY OF PREDICTIVE MODELS FOR CUSTOMER
RETENTION,
FINANCIAL
FORECASTING,
AND
INVENTORY
OPTIMIZATION.
International
Interdisciplinary Business Economics Advancement
Journal, 5(11), 21-32.
Bhattacharjee, B., Mou, S. N., Hossain, M. S.,
Rahman, M. K., Hassan, M. M., Rahman, N., ... &
Haque, M. S. U. (2024). MACHINE LEARNING FOR
COST ESTIMATION AND FORECASTING IN BANKING:
A COMPARATIVE ANALYSIS OF ALGORITHMS.
Frontline Marketing, Management and Economics
Journal, 4(12), 66-83.
Rahman, A., Iqbal, A., Ahmed, E., & Ontor, M. R. H.
(2024). PRIVACY-PRESERVING MACHINE LEARNING:
TECHNIQUES,
CHALLENGES,
AND
FUTURE
DIRECTIONS IN SAFEGUARDING PERSONAL DATA
MANAGEMENT. Frontline Marketing, Management
and Economics Journal, 4(12), 84-106.
Rahman, M. M., Akhi, S. S., Hossain, S., Ayub, M. I.,
Siddique, M. T., Nath, A., ... & Hassan, M. M. (2024).
EVALUATING MACHINE LEARNING MODELS FOR
OPTIMAL CUSTOMER SEGMENTATION IN BANKING:
A COMPARATIVE STUDY. The American Journal of
The American Journal of Applied Sciences
16
https://www.theamericanjournals.com/index.php/tajas
The American Journal of Applied Sciences
Engineering and Technology, 6(12), 68-83.
Das, P., Pervin, T., Bhattacharjee, B., Karim, M. R.,
Sultana, N., Khan, M. S., ... & Kamruzzaman, F. N. U.
(2024). OPTIMIZING REAL-TIME DYNAMIC PRICING
STRATEGIES IN RETAIL AND E-COMMERCE USING
MACHINE LEARNING MODELS. The American
Journal of Engineering and Technology, 6(12), 163-
177.
Al Mamun, A., Hossain, M. S., Rishad, S. S. I.,
Rahman, M. M., Shakil, F., Choudhury, M. Z. M. E.,
... & Sultana, S. (2024). MACHINE LEARNING FOR
STOCK MARKET SECURITY MEASUREMENT: A
COMPARATIVE
ANALYSIS
OF
SUPERVISED,
UNSUPERVISED, AND DEEP LEARNING MODELS. The
American Journal of Engineering and Technology,
6(11), 63-76.
