CRED-1: Open-source domain credibility dataset

CRED-1: Open-Source Domain Credibility Dataset

CRED-1 is a valuable open-source dataset designed for developers and researchers focused on evaluating the credibility of news sources. This dataset comprises 2673 domains, meticulously curated to facilitate the development and testing of AI models and systems that assess information reliability. Its integration with a TypeScript library, command-line interface (CLI), and MCP server makes it readily accessible and adaptable for various development workflows.

What CRED-1 Does

CRED-1 provides a structured collection of domain names, each associated with an assessment of its credibility. This allows developers to train and validate AI models that can predict or classify the reliability of news domains. By offering a standardized dataset, CRED-1 aims to foster reproducible research and accelerate the development of more robust and accurate systems for identifying trustworthy information online.

Key Features

• Comprehensive Dataset: Includes 2673 domains for in-depth analysis of news source reliability.
• Open-Source: Freely available for use, modification, and distribution, promoting community collaboration.
• Developer-Friendly Tools: Comes with a TypeScript library for programmatic access and a CLI for easy integration into existing pipelines.
• MCP Server Integration: Designed to work seamlessly with MCP servers, enabling distributed and scalable credibility assessment.
• Focus on Domain Credibility: Specifically targets the evaluation of domain-level reliability, a crucial aspect of information verification.

Who CRED-1 is For

CRED-1 is primarily intended for AI developers working on natural language processing (NLP), machine learning (ML), and information retrieval tasks. This includes researchers developing algorithms for fake news detection, content moderation platforms, and systems that require reliable data sources. Data scientists and software engineers building applications that need to filter or rank information based on source credibility will find CRED-1 to be a foundational resource. Its open-source nature and developer-centric tools make it ideal for individuals and teams looking to contribute to or leverage advancements in digital information integrity.