DescribeML: VS Code Plugin for Structured Dataset Documentation

1 Visual Studio Code language plug-in designed for the creation of structured machine learning dataset documentation. It provides a framework for recording metadata across three core categories: composition, provenance, and social concerns.

Use Cases

• Standardize dataset metadata using the composition and provenance fields
• Track data lineage by recording source information in the provenance section
• Identify ethical risks by documenting social concerns within the structured schema

Strengths

• Visual Studio Code language plug-in format
• 3 core documentation categories: composition, provenance, and social concerns
• Structured metadata output for dataset descriptions