User Sentiment Prediction and Analysis for Payment App Reviews Using Supervised and Unsupervised Machine Learning Approaches

Abstract

Businesses must be aware of customer sentiment in order to provide the best customer service. Instead of using cash or a credit card, a user can use a payment app on a mobile device to pay for a variety of services and digital or physical goods, which is becoming increasingly popular around the world. The goal of this study is to evaluate and predict user sentiment for payment apps using supervised and unsupervised machine learning (ML) approaches. For the study's data, Google Play Store reviews of the PayPal and Google Pay apps were gathered. Following cleaning, the filtered summary sentences were assessed for positive, neutral, or negative feelings using two unsupervised and five supervised machine learning approaches. According to the findings of the current study, the majority of customer reviews for payment apps were positive, with the average number of words with negative sentiment being higher. Furthermore, recent research found that, while all ML approaches can correctly classify review text into sentiment classes, logistic regression outperforms them in terms of accuracy.