All writing
RText AnalysisData ScienceNLP

Text Analysis of MyAirtel App Google Play Store Reviews in R

Overview

This analysis examines Google Play Store reviews for the MyAirtel mobile application — covering online recharge, bill payments, and money transfer services.

Data Source

Dataset scraped from Google Play Store using Beautiful Soup (Python), loaded into R for analysis. The dataset contains 41,819 reviews across 8 columns including review content, ratings, engagement metrics, version information, and company responses.

Key Findings

Most Common Words

Stop words and generic terms (app, Airtel) filtered to reveal meaningful vocabulary patterns in user feedback.

Sentiment Analysis

Positive words: excellent, amazing, awesome, perfect, wonderful, convenient, easy, fast

Negative words: otp, downloaded, bad, downloading, download, phone

Critical insight: "otp" (One Time Password) gets a very low average rating (1.37/5) — suggesting login and signup problems significantly impact user satisfaction.

Sentiment Trends Over Time

Analysis revealed encouraging patterns: negative sentiment declined while trust sentiment increased, indicating improving app performance perception over the observation period.

Word Relationships (Bigrams)

Network visualisation showed "app" frequently preceded by positive descriptors, while "data" commonly paired with terms like "purchased," "bonus," and "active."

Negation Impact

Phrases "not good" and "not happy" were identified as major causes of sentiment misclassification — demonstrating the importance of contextual analysis in NLP work.