# Customer Experience Analytics for Ethiopian Top 3 Banking Apps
## π Project Overview
This project analyzes user reviews from the Google Play Store for Ethiopia's top three banking applications:
- Commercial Bank of Ethiopia
- Bank of Abyssinia (BoA Mobile)
- Dashen Bank
The analysis provides actionable insights on customer satisfaction, key pain points, and feature enhancement opportunities to help banks improve their digital user experience.
## π― Key Objectives
- **Data Collection**: Scrape 400+ user reviews per banking app from Google Play Store
- **Text Processing**:
- Clean and preprocess reviews
- Remove duplicates
- Normalize dates
- Translate non-English reviews
- **Sentiment Analysis**:
- Implement DistilBERT for sentiment classification
- Handle multiple languages including Amharic
- **Theme Extraction**:
- Identify key topics and themes
- Group feedback into actionable categories
- **Data Storage**: Store processed data in Oracle database
- **Reporting**: Generate visualizations and Excel exports
## π Project Structure
```
βββ .github/ # GitHub workflows (CI/CD)
βββ .vscode/ # VSCode settings
βββ data/ # Raw and processed datasets (gitignored)
βββ notebooks/ # Jupyter notebooks for analysis
β βββ 1_scrape_and_explore.ipynb
β βββ 2_preprocessing_and_cleaning.ipynb
β βββ 3_sentiment_and_keywords.ipynb
β βββ 4_theming_and_reporting.ipynb
βββ outputs/ # Generated reports and plots
βββ scripts/ # Modular pipeline scripts
β βββ run_pipeline.py # End-to-end pipeline
β βββ scrape_reviews.py # Review collection
β βββ preprocess_reviews.py # Data cleaning
β βββ analyze_sentiment.py # Sentiment analysis
β βββ extract_keywords.py # Keyword extraction
β βββ assign_themes.py # Theme assignment
β βββ generate_report.py # Report generation
βββ src/ # Core modules
β βββ scraper.py # Google Play scra β¦