Fact Checker

This project is an API server for performing fact-checking on given texts. When a text is received, it is sent to a queue as a fact-checking task. The fact-checking results are classified into four categories: accurate, inaccurate, false, and indeterminate. The results can be retrieved using the task ID.

Setup and Run

Prerequisites

• Python 3.8 or higher
• AWS account
• AWS CLI configured with your credentials

Installation

1. Clone the repository:

   git clone https://github.com/HikaruEgashira/fact-checker.git
   cd fact-checker

2. Install the required dependencies:

   pip install -r requirements.txt
   cp config/pre-* .git/hooks

3. Deploy the infrastructure using Terraform:

   cd infra/production
   terraform init
   terraform apply

Running the API Server

To run the API server locally, use the following command:

python src/app.py

API Documentation

Endpoints

POST /fact-check

• Description: Submit a prompt for fact-checking.
• Request Body:

  {
    "prompt": "The earth is round."
  }

• Response:

  {
    "id": "unique-task-id"
  }

GET /fact-check/{id}

• Description: Retrieve the fact-checking result for a given task ID.
• Response:

  {
    "id": "unique-task-id",
    "result": "accurate" | "almost accurate" | "misleading" | "inaccurate" | "unsupported" | "incorrect" | "false" | "indeterminate" | "out of scope"
  }

CLI Usage

Submit a Fact-Checking Request

To submit a fact-checking request via the CLI, use the following command:

python src/app.py submit --prompt "The earth is round."

Check Task Status

To check the state of a fact-checking task via the CLI, use the following command:

python src/app.py state --id "unique-task-id"