Yachay.ai is an open-source machine learning community specializing in geolocation detection from natural language text. The project has collected, cleaned, and annotated decades of textual data from diverse sources including traditional media (e.g., New York Times), social media (Twitter, Reddit), messenger channels, tech blogs, GitHub profiles and issues, the dark web, and legal proceedings, as well as government regulatory and legislative documents. This curated dataset is used to train a state-of-the-art geolocation detection model, which has been released as an open-source tool. The community aims to advance research at the intersection of geospatial analysis and natural language processing (NLP), and actively seeks developers to contribute to and improve the project.
Key Features
- Curated Multisource Dataset: Cleaned and annotated text data from traditional and social media, dark web, legal, and governmental sources.
- State-of-the-Art Geolocation Model: Open-source geolocation detection tool built on deep learning.
- Community Infrastructure: GitHub repository, Kaggle competitions, and Discord channel for collaboration.
Who It's For
- Developers and researchers interested in geospatial NLP.
- Deep learning practitioners focused on location inference from text.
- Contributors seeking open-source projects in geolocation and NLP.
Key Benefits
- Open-source and freely available
- State-of-the-art geolocation detection model
- Comprehensive annotated dataset from diverse sources
- Active community with GitHub, Discord, and Kaggle engagement
