TriShield is a machine learning-based web application that classifies text as AI-generated or human-written. Its primary goal is to improve the fairness of AI text detection by reducing false positives for non-native English speakers, whose simple writing styles are often mistakenly identified as AI-generated. The project combines a lightweight web interface with a machine learning backend for accurate and accessible text classification.
The dataset is not included in this repository due to its size.
Download it from: https://drive.google.com/drive/folders/1B0orRwsEs57JzgRmXsSmzHQz3IcxRHFU?usp=drive_link
The Colab notebook is currently configured to load the dataset from Google Drive using the original file paths. If you are running the notebook on your own Google Colab account, you will need to:
- Download or copy the dataset to your own Google Drive.
- Mount your Google Drive in Colab.
- Update the dataset paths in the notebook to match the location where you stored the files.