In today's digital era, misinformation poses a significant threat, influencing decisions and shaping global opinions with often inaccurate or deceptive content. Nivara is a solution designed to combat this issue by creating a reliable platform that leverages the combined power of academic expertise and AI-driven models to provide verified, fact-checked information.
Nivara is a misinformation detection and verification system. It offers a platform where academic scholars verify the accuracy of articles and other media. By utilizing Natural Language Processing (NLP) and Artificial Intelligence (AI), the app also automates the initial analysis of content, flagging it as potentially truthful or misleading. This multi-layered approach aims to ensure the reliability of information consumed by users.
Nivara integrates subject-matter expertise as a core part of its fact-checking process. Qualified scholars review articles based on their expertise, assessing the truthfulness of the content.
- Scholars are matched to articles based on their domain knowledge (e.g., medical articles are reviewed by doctors or medical researchers).
- Verified experts provide a confidence score after reviewing an article, which is displayed alongside the content.
Nivara employs state-of-the-art NLP models to assist with article evaluation. These models perform an initial pass to flag questionable content:
- NLP algorithms assess linguistic cues and cross-reference available data to detect bias, inconsistencies, or misinformation.
- AI models assign a probability score that indicates whether the content is likely to be true or false, which is then reviewed by scholars for a more detailed examination.
Scholars are incentivized to contribute by being rewarded for accurate and thorough reviews. Features include:
- Reward Mechanism: Scholars earn points or credits based on the quality and frequency of their reviews.
- Performance-Based Incentives: A reputation system ensures that scholars with a history of accurate reviews are prioritized for more content and higher rewards.
Nivara provides an intuitive, clean interface where users can:
- Instantly see the credibility score of articles.
- Access detailed in-depth analysis (via a premium subscription).
- Interact with visualizations that show how the reliability of an article changes over time as more reviews and analysis are conducted.
Nivara is built for future expansion, allowing for additional media types to be verified:
- Images and Videos: Upcoming features will enable image and video fact-checking, addressing visual misinformation and deepfakes.
- The system is designed to handle an increasing number of users, reviews, and content types without sacrificing performance.
To run this project locally, follow these steps:
-
Clone the repository:
git clone https://github.com/Devonkedev/Niva.git
-
Navigate into the project directory:
cd Niva -
Install dependencies for both frontend and backend:
# Install server dependencies cd backend npm install # Install client dependencies cd ../frontend npm install
-
Run the application:
# Run server cd backend npm start # Run client cd ../frontend npm start
-
Visit
http://localhost:3000in your browser to access the app.
Contributions are welcome! If you'd like to contribute to the project:
-
Fork the repository.
-
Create a new branch with your feature or bug fix:
git checkout -b feature-name
-
Commit your changes:
git commit -m "Your message" -
Push to your branch:
git push origin feature-name
-
Open a Pull Request, detailing your changes.
This project is licensed under the MIT License. See the LICENSE file for details.