NLP • Three.js • FastAPI • React

3D Word Cloud Visualization

An interactive web app that analyzes article content in real time and transforms the most important topics into an immersive 3D word cloud for visual exploration.

GitHub

Project Overview

This project takes a news article URL, scrapes the content, extracts key terms using NLP techniques, and turns the result into a responsive 3D visualization. The experience combines backend text analysis with a frontend that is intentionally visual, interactive, and exploratory.

How It Works

Accepts a news article URL as input from the user interface.
Fetches and extracts the article content on the backend.
Processes the text using tokenization, stop-word removal, and TF-IDF.
Maps topic importance to spatial position, size, and color in a 3D scene.
Lets users rotate, zoom, and hover to explore the output interactively.

Highlights

NLP Pipeline

Uses TF-IDF and text preprocessing to surface meaningful article keywords.

URL-Based Analysis

Processes real article links instead of relying on static datasets.

3D Visualization

Uses React Three Fiber and Three.js for an immersive topic layout.

Fast Feedback

Designed for real-time interaction rather than offline-only analysis.

Technologies Used

Frontend: React 18, TypeScript, Three.js, React Three Fiber, Vite
Backend: FastAPI, BeautifulSoup4, scikit-learn, NLTK
Deployment: Vercel frontend and Render backend workflow
User Experience: Rotatable, zoomable, hover-based visual exploration