Roundabout - Hackathon Project

Overview

This project is a Machine Learning / NLP pipeline to evaluate the quality and relevancy of Google location reviews.
It includes preprocessing, feature extraction, model training, policy enforcement, and a simple UI for predictions.

---

START UP

1. Clone the repository

git clone https://github.com/xyanjun02/roundabout.git
cd roundabout

2. Set up the environment

# Create and activate the conda environment
conda env create -f environment.yml
conda activate roundabout

3. Install additional dependencies (if needed)

pip install -r requirements.txt

---

Folder Structure

tiktok-techjam/
├─ README.md
├─ .gitignore
├─ data/
│  ├─ raw/            # Original unprocessed datasets
│  └─ processed/      # Cleaned and preprocessed data
├─ scripts/           # Standalone scripts (e.g., download data)
├─ src/
│  ├─ preprocess/     # Data cleaning and text preprocessing
│  ├─ features/       # Feature extraction modules
│  ├─ policies/       # Rule-based or policy enforcement modules
│  ├─ llm/            # ML/NLP models (training and inference)
│  └─ utils/          # Utility functions (logging, configs, etc.)
├─ ui/                # Frontend UI (Streamlit / FastAPI)
├─ outputs/
│  ├─ models/         # Saved trained models
│  └─ predictions/    # Output predictions / results
├─ experiments/       # Jupyter notebooks or experiments
└─ requirements.txt   # Optional pip requirements file

---

Running the Project

---

Notes

• Ensure the roundabout conda environment is active before running scripts.
• Outputs (models, predictions) are saved in outputs/.
• Raw and processed data are not tracked in git (.gitignore) for team convenience.

---

Team / Contributors

• Steve Chia
• Xie Yanjun
• Tong Jia Jun
• Venice Phua
• Lee Sze Ying