dataset

Version: v26.0610
Namespace: https://behaverse.org/schemas/dataset

A metadata schema for cognitive science and neuroscience datasets with mappings to Schema.org, DataCite, BIDS, and other standards

Properties

This schema defines 45 properties for describing dataset metadata.

Core Metadata

Essential fields for dataset identification

Property	Type	Requirement	Description
@type	schema:Dataset	optional	JSON-LD node type (rdf:type) for schema.org / Google Dataset Search discoverabil...
name	string	required	Short URL-friendly identifier
pretty_name	string	recommended	Human-readable display title
description	string	required	Comprehensive dataset description
version	string	recommended	Dataset version (semantic versioning)
license	enum	required	License identifier in SPDX format
url	string	recommended	Dataset homepage or landing page
doi	string	recommended	Digital Object Identifier
keywords	array	recommended	Keywords for discovery
language	array	optional	ISO 639-1 language codes

Dates

Temporal metadata

Property	Type	Requirement	Description
date_created	string	recommended	Date dataset was originally created (ISO 8601)
date_published	string	recommended	Date first published (ISO 8601)
date_modified	string	optional	Date of last modification (ISO 8601)
date_added	string	required	Date added to catalog (ISO 8601, internal)
last_verified	string	optional	Date metadata was last verified (ISO 8601, internal)

People & Organizations

Creator and contributor information

Property	Type	Requirement	Description
creator	array	recommended	Primary dataset creators/authors
curator	array	optional	Dataset curators responsible for maintenance and quality

Creator Object

Each creator in the creator array is an object with the following properties:

Property	Type	Requirement	Description
name	string	required	Creator full name
email	string	optional	Creator email address (format: email)
orcid	string	optional	ORCID identifier (format: 0000-0000-0000-0000)
affiliation	string	optional	Institutional affiliation

Example:

{
  "creator": [
  {
    "name": "Jane Researcher",
    "email": "[email protected]",
    "orcid": "0000-0001-2345-6789",
    "affiliation": "University Psychology Department"
  }
]
}

Curator Object

Each curator in the curator array is an object with the following properties:

Property	Type	Requirement	Description
name	string	required	Curator full name
email	string	optional	Curator email address (format: email)
orcid	string	optional	ORCID identifier (format: 0000-0000-0000-0000)
affiliation	string	optional	Institutional affiliation

Citations & References

Published papers and references

Property	Type	Requirement	Description
citation	array	recommended	Published papers or references

Citation Object

Each citation in the citation array is an object with the following properties:

Property	Type	Requirement	Description
type	string	optional	Citation type
doi	string	optional	DOI of cited paper
url	string	optional	URL of cited resource
text	string	optional	Full citation text
arxiv_id	string	optional	ArXiv preprint ID

Population & Sample

Aggregate-level population demographics and coverage

Property	Type	Requirement	Description
sample_size	integer	required	Total number of participants
age_range	array	recommended	[min, max] age in years
age_mean	number	optional	Mean participant age
age_std	number	optional	Standard deviation of age
sex_distribution	object	recommended	Participant counts by sex
age_category	array	optional	Age group classification
population_category	enum	optional	Population type (clinical vs healthy)
inclusion_criteria	array	optional	Participant inclusion criteria
exclusion_criteria	array	optional	Participant exclusion criteria
spatial_coverage	string	optional	Geographic location/region
temporal_coverage	string	optional	Time period data covers

Sex Distribution Object

Each sex_distribution in the sex_distribution array is an object with the following properties:

Property	Type	Requirement	Description
female	integer	optional	Number of female participants
male	integer	optional	Number of male participants
other	integer	optional	Number of participants with other gender identity
not_reported	integer	optional	Number of participants who did not report gender

Example:

{
  "sex_distribution": {
  "female": 52,
  "male": 45,
  "other": 2,
  "not_reported": 1
}
}

Data Modalities & Measurement

Types of data collected and measurement techniques

Property	Type	Requirement	Description
measurement_technique	array	recommended	Measurement techniques used with optional detailed specifications
constructs_measured	array	optional	Cognitive or psychological constructs measured in the dataset

Measurement Technique Object

Each measurement_technique in the measurement_technique array is an object with the following properties:

Property	Type	Requirement	Description
type	string	optional	High-level category of measurement
technique	string	required	Specific measurement technique
channels	integer	optional	Number of channels/electrodes (for EEG, MEG)
sampling_rate	number	optional	Sampling rate in Hz (for EEG, MEG, physiological)
reference	string	optional	Reference type (for EEG, MEG)
manufacturer	string	optional	Equipment manufacturer/system
field_strength	number	optional	Magnetic field strength in Tesla (for MRI)
tr	number	optional	Repetition time in ms (for fMRI)
te	number	optional	Echo time in ms (for MRI)
details	string	optional	Additional technique-specific details
response_type	array	optional	Types of behavior responses (applies when technique is behavior)
format	string	optional	File format for this specific measurement (e.g., edf, bdf, nii, csv)
granularity	string	optional	Granularity of the measurement (e.g., per trial, per subject, per session)

Example:

{
  "measurement_technique": [
  {
    "type": "electrophysiology",
    "technique": "EEG",
    "channels": 64,
    "sampling_rate": 512,
    "reference": "average",
    "manufacturer": "BioSemi",
    "granularity": "event-data"
  }
]
}

Activities & Paradigms

Cognitive tasks and experimental paradigms

Property	Type	Requirement	Description
activity	array	recommended	Activities or tasks performed by participants with associated measurements

Activity Object

Each activity in the activity array is an object with the following properties:

Property	Type	Requirement	Description
name	string	required	Activity/task name
type	string	optional	Activity type/category
measurements	array	optional	List of measurement techniques collected during this activity (reference by technique name)
trials	integer	optional	Number of trials
duration	number	optional	Typical duration in minutes
conditions	array	optional	Experimental conditions
measures	array	optional	Primary dependent variables
constructs	array	optional	Constructs measured

Example:

{
  "activity": [
  {
    "name": "N-Back",
    "type": "task",
    "measurements": [
      "EEG",
      "eye-tracking",
      "response-device"
    ],
    "trials": 150,
    "duration": 20,
    "conditions": [
      "0-back",
      "2-back"
    ],
    "measures": [
      "d_prime",
      "reaction_time"
    ],
    "constructs": [
      "working memory"
    ]
  }
]
}

Study Design

Study design and methodology

Property	Type	Requirement	Description
study_design_type	enum	recommended	Study design category
intervention_type	array	optional	Type(s) of intervention (applies when study_design_type is intervention)
session_count	integer	optional	Number of experimental sessions per participant
session_description	string	optional	Brief description of session structure (if multi-session)

Data Files & Distribution

File formats and access information

Property	Type	Requirement	Description
data_formats	array	recommended	File formats (csv, json, edf, etc.)
data_size_gb	number	optional	Total dataset size (GB)
data_structure	string	optional	Organization of data files
download_url	string	optional	Direct URL to download dataset files (e.g., .zip archive or direct file download...
access_url	string	optional	URL to dataset landing page with documentation, access instructions, and metadat...
access_conditions	object	optional	Access requirements and restrictions for the dataset

Access Conditions Object

Each access_conditions in the access_conditions array is an object with the following properties:

Property	Type	Requirement	Description
is_free	boolean	optional	Whether the dataset is freely accessible without cost
requirements	string	optional	Access restrictions or requirements (e.g., registration, data use agreement)

Example:

{
  "access_conditions": {
  "is_free": true,
  "requirements": "Publicly available"
}
}

Ethics

Ethical approval and data quality

Property	Type	Requirement	Description
ethical_approval	object	optional	Ethical approval and IRB information

Ethical Approval Object

Each ethical_approval in the ethical_approval array is an object with the following properties:

Property	Type	Requirement	Description
obtained	boolean	optional	Whether ethics approval was obtained
institution	string	optional	IRB or ethics board institution
protocol	string	optional	Protocol or approval number

Example:

{
  "ethical_approval": {
  "obtained": true,
  "institution": "University IRB",
  "protocol": "IRB-2024-001"
}
}

HuggingFace-Specific

ML dataset metadata

Property	Type	Requirement	Description
size_category	enum	optional	HF size bucket
task_categories	array	optional	ML task types

Usage

See the examples for practical usage patterns.

Version History

The current version of dataset is v26.0610.

Older versions are available in the dataset/versions/ directory.