This project performs sentiment analysis and topic modelling on review data using Google Cloud Natural Language API and Vertex AI for TF-IDF and LDA.

• Project Presentation: Google Slides

• Python 3.8+
• Google Cloud account with:
  • Cloud Storage API enabled
  • Natural Language API enabled
  • Vertex AI API enabled
• Service account with appropriate permissions

Create and activate a virtual environment:

# Create virtual environment
python -m venv venv

# Activate on Linux/Mac
source venv/bin/activate

# Activate on Windows
# venv\Scripts\activate

pip install -r requirements.txt

1. Download your service account JSON key from Google Cloud Console:

   • Go to IAM & Admin → Service Accounts
   • Create or select a service account with the required permissions
   • Keys tab → Add Key → Create new key → JSON format

2. Save the downloaded file as service-account-key.json in the project root

   • This file is automatically gitignored for security
   • The notebook expects this exact filename and location

• bucket_name: Your GCS bucket name (currently ubc-bolt-case)
• project: Your Google Cloud project ID (currently sent-analysis-452609)

1. Open reviews_notebook.ipynb in Jupyter or VS Code
2. Run the cells sequentially:
   • Section 1: Sentiment Analysis - Computes sentiment scores for reviews
   • Section 2: Topic Modeling - Classifies reviews into categories
   • Section 3: Data Visualization - Creates plots and insights

• The .gitignore file is configured to exclude all *.json credential files
• Always use environment variables or local files excluded from git
• Rotate your keys immediately if they are accidentally exposed

The notebook expects:

• Input: reviews.csv in your GCS bucket
• Output: reviews_final_model_results.csv saved to GCS

See LICENSE file for details.

Contributor: Mikail Durrani, Jacky Zhong and me (ofc)