Sentiment classification based on machine learning approaches in flipkart product reviews

Abstract

E-commerce platforms and sellers constantly seek consumer feedback in the current digital retail environment, particularly regarding the customers' experiences with the product. The volume of customer evaluations and product comments posted online has significantly increased as a result of this new trend, which reflects a rising preference for online buying via the well-known Indian e-commerce site Flipkart (https://www.flipkart.com). The opinions and experiences of other customers are frequently relied upon by prospective buyers, therefore the sentiments conveyed in these evaluations have a considerable impact on their decisions. We focused our investigation on iPhone product reviews. The Bag-of-Words methodology was used to convert iPhone product reviews into numerical representations to start the process. Additionally, a word cloud was used to show word frequency throughout the evaluations to graphically demonstrate the text-based data. The investigation was then carried out using a range of machine learning techniques, including decision trees and logistic regression. Impressively, when examining the dataset made up of reviews of iPhone products and the Flipkart website, both machine learning models displayed outstanding accuracy. Notably, With a 99% exceptional accuracy rate, the Decision Tree model surpassed the Logistic Regression model, which was only 92% accurate at the time.