Back to Gap Map Model components / Data hubs & catalogs

UCI Machine Learning Repository

closed / Overall score: n/a

The UCI Machine Learning Repository is the University of California, Irvine's collection of datasets donated for machine learning research, kept since 1987. Librarians approve each donation, which then receives a DOI and a CC BY 4.0 license, and datasets from elsewhere can be linked. A public JSON API and the ucimlrepo Python package fetch a dataset by ID or name into pandas with its variable descriptions.

Scored as the hosted repository. The ucimlrepo client is open source, and the repository's public GitHub account otherwise holds an issue tracker.

Openness

1 high confidence

1.0

license: proprietary(hosted service
source: closed(hosted service

The repository's website and API are run by UC Irvine and their code is not published. The ucimlrepo package that fetches from it is open source, but it is a client and does not make the repository itself runnable elsewhere.

https://archive.ics.uci.edu/about recorded 2026-10-07
About page: librarians, NSF funding and "The current version of the website was released in 2023"; no source link.
https://raw.githubusercontent.com/uci-ml-repo/ucimlrepo/HEAD/README.md recorded 2026-10-07
"Package to easily import datasets from the UC Irvine Machine Learning Repository into scripts and notebooks." A client of the hosted service.

Adoption

not assessed

Dataset pages show running view and citation counts, but these are cumulative page views rather than downloads, and the repository publishes no monthly or total usage figure. No reading was recorded.

https://archive.ics.uci.edu/dataset/53/iris recorded 2026-10-07
Iris dataset page: "352 citations 1522782 views", DOI 10.24432/C56C76, read 2026-10-07.

Capability

2 medium confidence

2.0

UCI hosts each donated dataset, gives it a DOI and serves it by ID through a public API that the ucimlrepo package loads into pandas. A dataset is updated in place under the same record rather than kept as numbered versions, which PhysioNet does.

https://archive.ics.uci.edu/api/dataset?id=53 recorded 2026-10-07
JSON record for Iris: "uci_id":53, "data_url":"https://archive.ics.uci.edu/static/public/53/data.csv", "dataset_doi":"10.24432/C56C76", "last_updated":"Tue Sep 12 2023", variables and feature types; no version field.
https://archive.ics.uci.edu/contribute/donation recorded 2026-10-07
"Datasets approved to be in the repository will be assigned a Digital Object Identifier (DOI)"; "Datasets will be licensed under a Creative Commons Attribution 4.0 International license".
https://raw.githubusercontent.com/uci-ml-repo/ucimlrepo/HEAD/README.md recorded 2026-10-07
"fetch_ucirepo(id=45)"; "Provide either a dataset ID or name"; returns "dataset matrices as pandas dataframes" and metadata.

Verified 2026-10-07