README.md

metadata

title: Interactive Topic Modeler
emoji: 🧠
colorFrom: indigo
colorTo: purple
sdk: gradio
python_version: '3.10'
app_file: app.py
pinned: false
license: mit

Interactive Topic Modeler

This is a premium, lightweight, interactive topic modeling tool designed specifically for computational social science and humanities courses.

Features

• Flexible Data Input: Upload .csv, .xlsx, or .txt datasets directly. For tabular data, select any target text column for analysis.
• Multiple Algorithms:
  • LDA (Latent Dirichlet Allocation): Fast, traditional topic modeling running locally on CPU.
  • NMF (Non-Negative Matrix Factorization): Classic linear algebraic topic modeling.
  • BERTopic (API Mode): Utilizes advanced neural embeddings via Hugging Face Serverless API with your personal access token.
• Interactive Visualizations: View plotly bar charts displaying keywords and importance weights per topic.
• Download Labeled Dataset: Automatically outputs a downloadable version of the original dataset annotated with the assigned topics and document probabilities.