📰 Fake News Detection

A fact-checking application built with FastAPI, Google Gemini API, and SerpAPI. It accepts a claim, fetches live evidence from the web, and processes it for analysis. Gemini then evaluates the context to return a verdict with reasoning, confidence score, and supporting evidence.

Features

• Accepts a claim as input (via API or frontend).
• Fetches live web search results using SerpAPI.
• Cleans and normalizes the claim text.
• Passes evidence snippets into Gemini 1.5 Flash for structured reasoning.
• Returns a structured verdict:

Final Verdict: REAL / FAKE / UNKNOWN Score: Confidence (0–100) Reasoning: Explanation of why Evidence: Context snippets Warnings: Weak/contradictory info if any

---

Project Structure

.
├── main2.py                # FastAPI backend with fact-check pipeline  
├── requirements.txt       # Python dependencies  
├── landing_page/          # (Optional) UI assets / screenshots                                                                          ---
├── search_query_result/   # Stored snippets/screenshots from search  
└── README.md              # Documentation  

1. Clone the repository

git clone https://github.com/<your-username>/fake-news-detection.git
cd fake-news-detection

2. Create a virtual environment

python -m venv .venv

3. Activate the virtual environment

On Windows

.venv\Scripts\activate


On macOS/Linux

source .venv/bin/activate

4. Install dependencies

pip install -r requirements.txt

5. Create a .env file

GEMINI_API_KEY=your-gemini-api-key
SERPAPI_API_KEY=your-serpapi-api-key

6. Run the backend

uvicorn main2:app --reload

API will be live at:

http://127.0.0.1:8000

Example Usage

curl -X POST "http://127.0.0.1:8000/fact-check" \
-H "Content-Type: application/json" \
-d '{"claim": "The Eiffel Tower is in Paris"}'

Sample Response

{
 "verdict": "REAL",
 "score": "95",
 "reasoning": "Multiple sources confirm the Eiffel Tower is in Paris.",
 "evidence": "- The Eiffel Tower is a landmark in Paris, France.",
 "warnings": ""
}

API Endpoints

POST /fact-check → Takes a claim and returns a fact-check result

Tech Stack

-Python 3.10+

-FastAPI (backend framework)

-Google Gemini API (fact-check reasoning)

-SerpAPI (live web search results)

##  Frontend Setup (Angular)

### 1. Navigate to the frontend folder
cd fact-chex

2.Install dependencies

npm install

3. Run the Angular app

ng serve

The frontend will be live at:

http://localhost:4200

Backend Setup (FastAPI)

Run the backend

uvicorn main2:app --reload

The backend will be live at:

http://127.0.0.1:8000

Connecting Frontend & Backend

The Angular frontend (http://localhost:4200) sends requests to the FastAPI backend (http://127.0.0.1:8000/fact-check).

📹 Demonstration Video

▶️Watch Demo

---

📚 References

1. Shu, K., Sliva, A., Wang, S., Tang, J., & Liu, H. (2017). Fake News Detection on Social Media: A Data Mining Perspective. ACM SIGKDD Explorations Newsletter, 19(1), 22- 36. https://doi.org/10.1145/3137597.3137600

2. Zhou, X., & Zafarani, R. (2020). A Survey of Fake News: Fundamental Theories, Detection Methods, and Opportunities. ACM Computing Surveys, 53(5), 1–40. https://doi.org/10.1145/3395046

3. Google AI. Gemini LLM Documentation. https://ai.google.dev

🖼️ Assets / Screenshots