LLM text classification Shiny App (R)

Author

Kaizhong (Mike) Mu

Published

August 1, 2025

<> Code

Shiny App

Data Files

The app requires three data files:

  1. Original Data
  2. Human-Filled Data
  3. GPT-Filled Data

Shiny App Readme

The goal of this shiny app is to automatically report the comparison of AI vs Human performance in text classification for Private Equity (in Comparison: Human vs. GPT tab) and report the exploratory data analysis (in EDA: Human-filled Dataset tab).

This will enable you to no longer need to repeatedly write code/reports when trying various LLM models with human comparisons in the future. All you need to do is input three datasets: “Original Data” (100 entries randomly selected from the original dataset), “Human-Filled Data” (humans text classification), and “GPT-Filled Data” (AI text classification given prompts)

Presentation

Project Overview

The shiny app is designed to provide visualization for the health data science summer fellowship projects, supervised by Dr. Alyssa Bilinski. You could check the details of project from the presentation slides.

Direct Link to App

Back to top

Citation

BibTeX citation:
@online{(mike)_mu2025,
  author = {(Mike) Mu, Kaizhong},
  title = {LLM Text Classification {Shiny} {App} {(R)}},
  date = {2025-08-01},
  url = {https://kaizhongmu.github.io/project/dsp/2/},
  langid = {en}
}
For attribution, please cite this work as:
(Mike) Mu, Kaizhong. 2025. “LLM Text Classification Shiny App (R).” August 1, 2025. https://kaizhongmu.github.io/project/dsp/2/.