veritas

APS360 Project - Fake News Detection System

Overview

Veritas is a comprehensive fake news detection system that uses multiple machine learning approaches to classify news articles as real or fake. The project includes:

• Base Model: Traditional machine learning approach using TF-IDF and Logistic Regression
• RNN Model: Deep learning approach using Recurrent Neural Networks
• Dataset Processing: Automated pipeline to process multiple fake news datasets

Poetry Scripts

The project provides convenient Poetry scripts for common tasks:

Script	Command	Arguments	Description
generate-data	poetry run generate-data	None	Process and combine multiple fake news datasets into a unified dataset
train-rnn	poetry run train-rnn	--epochs <int>	Train the RNN model for fake news detection
train-base	poetry run train-base	None	Train the base model using TF-IDF and Logistic Regression
visualize-rnn	poetry run visualize-rnn	--history-path <str> [--output-path <str>]	Generate visualizations of RNN training results
visualize-base	poetry run visualize-base	None	Generate visualizations of base model performance

Command Line Arguments

train-rnn

• --epochs <int> (required): Number of epochs to train for

Example:

poetry run train-rnn --epochs 50

visualize-rnn

• --history-path <str> (required): Path to the training history JSON file
• --output-path <str> (optional): Path to save the output visualization. Defaults to "visualizations/training_history.png"

Example:

poetry run visualize-rnn --history-path history/training_history/model_128_0.001_0.0.json --output-path visualizations/my_training_results.png

Other Scripts

• generate-data: No arguments required
• train-base: No arguments required
• visualize-base: No arguments required

These scripts are defined in pyproject.toml and provide a simple interface to the project's main functionality.

Getting Started

This guide will help you set up the development environment for the veritas project.

Prerequisites

• Python 3.10 or higher
• Poetry (Python dependency management tool)

Installing Poetry

If you don't have Poetry installed, follow the official installation guide:

On macOS/Linux:

curl -sSL https://install.python-poetry.org | python3 -

On Windows:

(Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | py -

Verify installation:

poetry --version

Setting up the Project

1. Clone the repository (if you haven't already):

   git clone <repository-url>
   cd veritas

2. Install dependencies using Poetry:

   poetry install

3. Activate the Poetry virtual environment:

   poetry shell

4. Verify the environment is active:

   which python  # Should point to Poetry's virtual environment
   pip list      # Should show all project dependencies

Code Linting with Black

This project uses Black for automatic code formatting to maintain consistent code style.

Installing Black

1. Install Black (if not already installed):

   poetry add --group dev black

2. Verify installation:

   poetry run black --version

Using Black with VSCode Extension

The easiest way to use Black is with the VSCode extension:

1. Install the Black Formatter extension in VSCode:

   • Open VSCode
   • Go to Extensions (Ctrl+Shift+X or Cmd+Shift+X)
   • Search for "Black Formatter"
   • Install the extension by Microsoft

2. Configure VSCode to use Black:

   • Open VSCode settings (Ctrl+, or Cmd+,)
   • Search for "python formatting provider"
   • Set it to "black"
   • Enable "Format on Save" for automatic formatting

3. Format your code:

   • Right-click in a Python file and select "Format Document"
   • Or use the keyboard shortcut: Shift+Alt+F (Windows/Linux) or Shift+Option+F (Mac)
   • Or enable "Format on Save" to automatically format when you save files

Command Line Usage (Optional)

If you prefer using Black from the command line:

# Format all Python files in the project
poetry run black .

# Format a specific file
poetry run black src/base_model/main.py

# Check what would be formatted without making changes
poetry run black --check .

Deactivating the Poetry Environment

When you're done working on the project:

exit  # If using poetry shell
# or
deactivate  # If the environment is still active

Troubleshooting

• If you get permission errors: Make sure you have write permissions in the project directory
• If Poetry fails to install dependencies: Try updating Poetry first: poetry self update
• If you need to recreate the environment: Delete the .venv folder and run poetry install again
• If Black formatting fails: Make sure you're in the Poetry environment and Black is installed
• If packages are missing: Run poetry install to ensure all dependencies are installed

Notes

• Always use poetry run or activate the Poetry environment before running scripts
• Poetry automatically manages virtual environments and dependencies
• The pyproject.toml file contains all project dependencies and configuration
• Run poetry run black . before committing to ensure consistent code formatting
• Poetry creates a .venv directory in your project folder for the virtual environment