Core Section

The core section contains essential identity and maintainer information for the dataset.

Required Fields

Field	Type	Description
`id`	string	Stable machine-friendly dataset ID (slug). Pattern: `^[a-zA-Z0-9._-]+$`
`version`	string	Dataset version label (e.g., “2025-02-01” or “1.0.0”)
`title`	string	Human-readable dataset name
`summary`	string	1-3 sentence description of the dataset
`maintainer`	string	Organization or individual responsible for the dataset
`contact`	string	Contact email or URL for questions

Optional Fields

Field	Type	Description
`created_at`	string (date-time)	When this dataset was first released
`last_modified_at`	string (date-time)	When this data card was last updated
`preferred_citation`	string	Preferred citation text for papers, reports, or downstream users
`citation_url`	string (uri)	URL with citation instructions or canonical citation entry
`doi`	string	Dataset DOI, if available. Pattern: `^10\.[0-9]{4,9}/[-._;()/:A-Za-z0-9]+$`

Example

{
  "core": {
    "id": "hausa-news-corpus",
    "version": "2025-01-15",
    "title": "Hausa News Corpus",
    "summary": "A curated collection of Hausa language news articles from Nigerian media outlets, covering politics, sports, and entertainment from 2020-2024.",
    "maintainer": "African Language Research Institute",
    "contact": "data@alri.example.org",
    "created_at": "2025-01-15T00:00:00Z",
    "last_modified_at": "2025-01-20T12:30:00Z",
    "preferred_citation": "ALRI Hausa News Corpus (2025). African Language Research Institute.",
    "doi": "10.5281/zenodo.12345678"
  }
}

Field Details

id

The id field must be a stable, machine-friendly identifier. It should:

Only contain letters, numbers, dots, hyphens, and underscores
Be unique within your organization or the registry
Not change once assigned

version

Version labels can follow any scheme you prefer:

Semantic versioning: 1.0.0, 2.1.3
Date-based: 2025-01-15, 2025-Q1
Custom: v1-alpha, release-2

contact

The contact field accepts either:

An email address: data@meetkai.ai
A URL: https://datapass.meetkai.ai/contact

This enables users to ask questions about the dataset.