File size: 3,875 Bytes
c83cd7e 7a5d5dd c83cd7e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 |
---
title: Falcon
emoji: 🔥
colorFrom: indigo
colorTo: yellow
sdk: docker
pinned: false
---
# Falcon: Fake News Analysis and Language Comprehension for Online Neutrality
**Falcon** is an advanced Gen AI-powered system designed to analyze, verify, and classify online claims to promote digital content neutrality. Leveraging state-of-the-art language models, it classifies user-submitted text, detects tone and intent, verifies facts using multiple data sources, and generates an informed verdict about claim credibility.
---
## Features
* **Claim Classification**: Classifies input text into categories like Factual Claim, Opinion, Irrelevant Talk, or Vague/Incomplete.
* **Tone and Intent Detection**: Analyzes the emotional tone and intent behind the claim (e.g., neutral, persuasive, humorous).
* **Fact Verification**: Verifies factual claims by searching reliable sources like Google (via Serper API) and Wikipedia.
* **Verdict Generation**: Produces a clear verdict on the claim’s truthfulness based on gathered evidence and reasoning.
* **Fallback Language Model**: Uses OpenAI’s GPT-4 as the primary LLM and automatically switches to **Deepseek** as a fallback model to ensure high availability and robustness.
* **Simple Web Interface**: Easy-to-use frontend using HTML, CSS, and JavaScript for seamless user interaction.
* **FastAPI Backend**: Fast, lightweight API server handling all processing with asynchronous endpoints.
---
## Tech Stack
* **Backend**: FastAPI (Python)
* **Frontend**: HTML, CSS, JavaScript
* **Language Models**: OpenAI GPT-4 (primary), Deepseek (fallback)
* **APIs**: Serper API (Google Search), Wikipedia API
* **Prompt Engineering**: Carefully crafted templates for accurate and reliable results
* **Deployment**: Docker containerization (optional)
---
## Project Structure
```
falcon/
├── backend/
│ ├── api/
│ │ ├── claims.py # Claim classification logic
│ │ ├── fact_check.py # Fact verification logic
│ │ ├── tone_intent.py # Tone and intent detection logic
│ ├── langchain_tools.py # LangChain prompt templates and LLM wrappers
│ ├── config.py # API keys and config variables
│ └── main.py # FastAPI app entry point
├── frontend/
│ ├── index.html # Main UI page
│ ├── assets/ # CSS, JS, images
├── Dockerfile # Docker container definition (optional)
└── README.md # This file
```
---
## Setup Instructions
### Prerequisites
* Python 3.9+
* FastAPI
* Uvicorn (ASGI server)
* OpenAI API key
* Serper API key
* Deepseek API key (for fallback)
### Installation
1. Clone the repository:
```bash
git clone https://github.com/yourusername/falcon.git
cd falcon
```
2. Create and activate a virtual environment:
```bash
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
```
3. Install dependencies:
```bash
pip install -r requirements.txt
```
4. Configure API keys in `backend/config.py`:
```python
OPENAI_API_KEY = "your_openai_api_key"
SERPER_API_KEY = "your_serper_api_key"
DEEPSEEK_API_KEY = "your_deepseek_api_key"
```
5. Run the FastAPI server:
```bash
uvicorn backend.main:app --port 8000
```
6. Open your browser and visit `http://localhost:8000` to use the Falcon app.
---
## Usage
* Enter a claim or statement in the input form.
* The system classifies the claim, detects tone and intent, performs fact verification if applicable, and returns a detailed verdict with supporting evidence.
* The fallback model Deepseek is automatically used if GPT-4 is unavailable, ensuring consistent performance.
|