JIITA, Vol.9 No.3 pp.1104-1121 (2025), DOI: 10.22664/ISITA.2025.9.3.1104
Thambusamy Velmurugan, and Mohandas Archana
Abstract. Nowadays the Customer reviews are becoming more and more important to businesses in the market because they have a
significant impact on consumer behavior and marketing strategy. Purchase decisions can be greatly influenced by the insightful
information provided by customer reviews regarding product performance and customer satisfaction. As a result, companies
leverage various techniques to analyze and interpret these reviews. A branch of Natural Language Processing (NLP) called sentiment
analysis is essential to comprehending the feelings conveyed in consumer reviews. By automating the classification of sentiments
such as positive, negative, or neutral, businesses can gain a deeper understanding of customer opinions and enhance their
marketing efforts. Customer reviews of a variety of musical products make up the musical instruments customer reviews dataset,
which was obtained from Kaggle. Several conventional preprocessing techniques, such as lowercase conversion, stopword removal,
stemming, punctuation and symbol removal, and lemmatization, were used to get the data ready for sentiment analysis.
Following cleaning, the data was converted to numerical form and divided into two sets: 80% for training and 20% for testing.
The sentiments were then divided into positive, negative, and neutral categories using a number of well-known machine
learning algorithms, such as Naive Bayes, Support Vector Machine (SVM), Random Forest, and Decision Tree
and NSRD (Naïve + Support Vector Machine + Random Forest + Decision Tree) proposed algorithm.
Accuracy, precision, recall, and F1-score metrics were used to assess these models’ performance, offering a thorough
examination of their efficacy in sentiment classification.
Keywords; Natural Language Processing, Naïve Bayes Algorithm, Random Forest Algorithm, Support Vector Machine
, Decision Tree Algorithm, NSRD (hybrid algorithm)
Fullpaper: