Sentiments Analysis on Amazon Product Reviews Using Supervised Machine Learning Algorithms

To determine the sentiment or emotion indicated in a text, such as a review, tweet, or piece of customer feedback, sentiment analysis, also known as opinion mining, uses computational linguistics, machine learning, and natural language processing techniques. Sentiment analysis seeks to determine whether the automatic sentiment expressed in the text is positive, negative, neutral, or occasionally more nuanced emotions like joy, sadness, rage, etc. The Apple iPhone 14 Pro Max (1 TB, Gold) evaluations from 160 Amazon consumers were used in this study. We used tokenization, stop word removal, stemming, and punctuation mark removal for preprocessing. Positive and negative word analysis is done using word clouds. We received favourable and unfavourable feedback about the Apple iPhone 14 Pro Max. Dynamic, good, improving, lovely, fantastic, great, etc., are positive evaluations. Additionally, unfavourable reviews are a costly, stuck gimmicky, gimmicky hesitant, heck of many complaints, bug bumpy and alarming, break dissatisfied, etc. We used Naive Bayes, Support Vector Machines (SVM), and Logistic Regression, among other machine learning approaches. We got 90%, 91%, and 94% accuracy for Logistic Regression, Support Vector Machines (SVM), and Naive Bayes, respectively.