This project analyzes user reviews from the Google Play Store to automatically identify, categorize, and prioritize customer complaints using NLP and unsupervised machine learning. It simulates a real-world feedback triage system to help product and support teams focus on high-impact issues efficiently.
- Automatically categorize 6,800+ app reviews into interpretable complaint topics
- Prioritize high-frustration issues using sentiment analysis
- Route each review to the appropriate business team (Tech, Product, Billing, Support)
- Simulate time savings by automating triage and complaint detection
- Topic Clustering with BERTopic: Identified top 5 complaint categories covering 91.3% of total reviews. -Sentiment Analysis: Flagged issues like "Notification Bugs" with 83%+ negative sentiment, guiding prioritization.
- Routing Simulation: Automatically assigned 100% of complaints to relevant teams.
- Time Savings Estimation: Automated analysis simulated 30+ hours of manual review saved.
A scalable and automated review analysis system that mimics real-world customer feedback management — helping product, billing, and tech teams prioritize and act on issues faster.