- Research
- Data
On May 7 all 22 controlled phenotype vocabularies were finished: 196 dimensions, 853 terms and 113 references. Each dimension is anchored to a published scale, from the Fitzpatrick skin types to the Heath-Carter somatotype and the Manning digit ratio.
A controlled vocabulary turns description into data. Every observation has to use a defined term, every term carries a written definition, and every dimension cites the scale it comes from.
The same day every group page gained a Data Depth score from 0 to 100 that shows how much evidence sits behind it. Dataset v3 on May 8 added the vocabularies and the coverage score for every group.
Under the hood
- Each dimension is a typed object, categorical, ordinal, numeric or structured, carrying its scale, the scale's citation and a definition for every allowed value.
- An observability block records whether a dimension can be read from a photograph and the minimum visible extent it needs, from the head alone to the full body. 167 of the 196 dimensions are photo-assessable, and the other 29 stay in the schema for completeness and are skipped during image analysis.
- Files are grouped by anatomical system, such as face and head, trunk and limbs, carry identifiers from standard anatomy ontologies, and are versioned with a documented bump policy.
- A code generator compiles each vocabulary file into four artifacts: the analysis prompt fragment, the data model, the aggregation function and a documentation page. Adding a dimension means editing one file.
- The Data Depth score combines four signals from a group's image observations: sample size, image quality, model confidence and source diversity. Dataset v3 publishes the vocabularies twice, as nested schemas and as a flat table with one row per dimension.
- controlled vocabularies
- 22
- dimensions
- 196
- defined terms
- 853
See it on the site
Sources and open data
- Controlled vocabulary Wikipedia
- EthnicErotic/phenotype-catalog Hugging Face
- Fitzpatrick (1988), The Validity and Practicality of Sun-Reactive Skin Types I Through VI Archives of Dermatology
- Heath and Carter (1967), A modified somatotype method American Journal of Physical Anthropology
- Manning et al. (1998), The ratio of 2nd to 4th digit length: a predictor of sperm numbers and concentrations of testosterone, luteinizing hormone and oestrogen Human Reproduction
- Uberon anatomy ontology OBO Foundry
- Foundational Model of Anatomy OBO Foundry
Related build notes
- May 7, 2026Open pipeline, methodology paper and a DOI: rebuild it yourselfThe code that builds the phenotype dataset is public under Apache 2.0, with a methodology paper and a Zenodo DOI, so anyone can rebuild and cite it.
- May 6, 202613,094 notable people and 5,668 photos read by a vision modelGroup pages gain Notable People linked to Wikipedia, and a vision model turns 5,668 reference photos into structured trait observations, each traceable.
- May 5, 2026The phenotype catalog goes open on Hugging FaceAll 484 groups, each with an original profile of 300 to 450 words, published as an open CC BY 4.0 dataset one load_dataset() call away.