Improve dataset card

README.md

@@ -74,126 +74,64 @@ dataset_info:
   dataset_size: 269611482
 ---
 
-# Dataset Card for ARTigo: Social Image Tagging
-
-## Table of Contents
-
-- [Table of Contents](#table-of-contents)
-- [Dataset Description](#dataset-description)
-  - [Dataset Summary](#dataset-summary)
-  - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
-- [Dataset Structure](#dataset-structure)
-  - [Data Instances](#data-instances)
-  - [Data Fields](#data-fields)
-  - [Data Splits](#data-splits)
-- [Additional Information](#additional-information)
-  - [Licensing Information](#licensing-information)
-  - [Citation Information](#citation-information)
-
-## Dataset Description
-
-- **Homepage:** https://www.artigo.org
-- **Repository:** https://github.com/arthist-lmu/artigo
-- **Data:** https://doi.org/10.5281/zenodo.8202331
-
-### Dataset Summary
-
-ARTigo (https://www.artigo.org/) is a Citizen Science project that has been jointly developed at the Institute for Art History and the Institute for Informatics at Ludwig Maximilian University of Munich since 2010. It enables participants to engage in the tagging of artworks, thus fostering knowledge accumulation and democratizing access to a traditionally elitist field. ARTigo is built as an interactive web application that offers Games With a Purpose: in them, players are presented with an image – and then challenged to communicate with one another using visual or textual annotations, *tags*, within a given time. Through this playful approach, the project aims to inspire greater appreciation for art and draw new audiences to museums and archives. It streamlines the discoverability of art-historical images, while promoting inclusivity, effective communication, and collaborative research practices. The project’s data are freely available to the wider research community for novel scientific investigations.
-
-### Supported Tasks and Leaderboards
-
-- `object-detection`: This dataset can be used to train models for object detection tasks on art-historical images.
-- `image-classification`: This dataset can also be used for image classification tasks by using only the tags and not the associated region information.
-
-## Dataset Structure
-
-This dataset has a single configuration.
-
-### Data Instances
-
-An example instance from this dataset:
-
-```python
-{
-  'id': 32254,
-  'hash_id': 'e34fa90bf4c73d20ac19b14fa615206e',
-  'titles': {
-    'id': [10893],
-    'name': ['Entwurf für ein zwölfteiliges Kartenspiel']
-  },
-  'creators': {
-    'id': [2391],
-    'name': ['Félix Vallotton']
-  },
-  'location': 'Lausanne',
-  'institution': 'Galerie du Chêne',
-  'source': {
-    'id': 2,
-    'name': 'Artemis',
-    'url': 'http://artemis.uni-muenchen.de/'
-  },
-  'path': 'https://api.artigo.org/media/e3/4f/e34fa90bf4c73d20ac19b14fa615206e.jpg',
-  'tags': {
-    'id': [6, 10, 13, ..., 206331],
-    'name': ['blau', 'feder', 'flügel', ..., 'herzober'],
-    'language': ['de', 'de', 'de', ..., 'de'],
-    'count': [16, 6, 6, ..., 1],
-    'regions': [None, None, None, ..., None]
-  },
-  'image': <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=381x600 at 0x7FEF3A415820>
-}
-```
+# ARTigo: Social Image Tagging
+
+60,633 digital reproductions of artworks with crowdsourced tags, from [ARTigo](https://www.artigo.org/) — a citizen-science project run since 2010 by the Institute of Art History and the Institute of Informatics at LMU Munich. Players are shown an image and type tags against a clock, scoring when a tag matches one their anonymous opponent gives or one recorded in an earlier session. The aggregate of those game rounds is this dataset. Built from the [v1.5 Zenodo deposit](https://doi.org/10.5281/zenodo.8202331) (1 August 2023), whose `data.jsonl` holds 60,981 records; the deposit is refreshed monthly, so current versions on Zenodo are larger than this build.
+
+Measured on that deposit: 270,024 distinct tags, 4.9M tag–image pairs, and just under 10M individual taggings, at a median of 61 distinct tags per image. Images come from nine sources, named in `source.name`: Artemis (26,402), Koeln (21,055), Rijksmuseum (5,633), Mkg-Hamburg (3,965), Kunsthalle-Karlsruhe (2,145), Amherst-College (1,365), and small contributions from Albertina, Stadt-Land-Bild and Streetart.
+
+## Fields
+
+| field | notes |
+| --- | --- |
+| `id`, `hash_id` | integer id; content hash used as the image filename |
+| `titles`, `creators` | lists of `{id, name}`; 8,769 distinct creator names |
+| `location`, `institution` | holding place and institution, where known |
+| `source` | `{id, name, url}` — one of the nine collections above |
+| `path` | URL of the image on the ARTigo API |
+| `tags` | list of `{id, name, language, count, regions}` |
+| `image` | the reproduction itself |
 
-### Data Fields
+Metadata coverage is partial: `creators` is empty on 12,182 records (20%), `institution` on 13,687 (22%), `location` on 17,795 (29%). Tags are unedited — the depositors deliberately left player spelling errors uncorrected rather than risk mangling specialist art-historical vocabulary.
 
-This dataset contains ten fields:
+> [!WARNING]
+> The repo still ships a legacy Python loading script, which `datasets` no longer runs (`RuntimeError: Dataset scripts are no longer supported`). Until it is converted to Parquet, work from the Zenodo deposit.
 
-- `id`: a unique identifier for the image;
-- `hash_id`: a unique identifier for the image based on its content (e.g., image hash);
-- `titles`: a list of titles associated with the image, with each title having the following key-value pairs:
-  - `id`: a unique identifier for the title;
-  - `name`: the name of the title;
-- `creators`: a list of creators associated with the image, with each creator having the following key-value pairs:
-  - `id`: a unique identifier for the creator;
-  - `name`: the name of the creator;
-- `location`: the location associated with the image;
-- `institution`: the institution that holds the image;
-- `source`: information about the source of the image, with the following key-value pairs:
-  - `id`: a unique identifier for the source;
-  - `name`: the name of the source;
-  - `url`: the URL of the source;
-- `path`: the path to the image file;
-- `tags`: a list of tags associated with the image, with each tag having the following key-value pairs:
-  - `id`: a unique identifier for the tag;
-  - `name`: the name of the tag;
-  - `language`: the language of the tag (if available);
-  - `count`: the number of times the tag has been applied to the image;
-  - `regions`: the regions of the image to which the tag can be applied (if available);
-- `image`: the image.
+## The tags are overwhelmingly German
 
-### Data Splits
+By tag–image pair: German 4,378,382 (89%), English 472,028 (9.6%), French 68,557 (1.4%). Anything trained on this without filtering learns a German tag vocabulary.
 
-This dataset doesn't provide any predefined train, validation or test splits. 
+`language` is also a weaker signal than it looks. It records which language version of ARTigo the tag was collected on, not the language of the tag: the data paper notes that "English tags created in the German version of ARTigo are labelled as German".
 
-## Additional Information
+## `regions` is almost always empty
 
-### Licensing Information
+Visual annotation — tagging a rectangle rather than the whole image — only started in 2022. In this deposit 521 records out of 60,981 (0.9%) carry any region at all, across 1,935 tag–image pairs. The `object-detection` task tag in the metadata is not supported at usable scale; in practice this is an image-tagging dataset. Where regions exist, the data paper describes them as normalised bounding boxes given by the x and y of the bottom-left corner plus width and height.
 
-[CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)
+## Use
 
-### Citation Information
+- Training or evaluating image tagging and text-to-image retrieval on art-historical images, in German.
+- `count` is a per-image agreement signal — head tags were given by many players, the long tail by one. It is the obvious weight or threshold for filtering noise.
+- Comparing player vocabulary against curatorial vocabulary, using `titles` and `creators` as the institutional description of the same object.
+
+## Licence
+
+The Zenodo record is CC BY-SA 4.0. That covers the deposit as a whole, including the aggregated annotations; the record does not break out rights per source institution for the reproductions themselves, which come from nine third-party collections. Check the source before redistributing images, and note that share-alike propagates to derivatives.
+
+## Citation
 
 ```
 @dataset{bry_et_al_artigo,
   author       = {Bry, François and
                   Kohle, Hubertus and
-                  Krefeld, Thomas and 
-                  Riepl, Christian and 
-                  Schneider, Stefanie and 
-                  Schön, Gerhard and 
+                  Krefeld, Thomas and
+                  Riepl, Christian and
+                  Schneider, Stefanie and
+                  Schön, Gerhard and
                   Schulz, Klaus},
   title        = {{ARTigo}: Social Image Tagging (Aggregated Data)},
   publisher    = {Zenodo},
   doi          = {10.5281/zenodo.8202331},
   url          = {https://doi.org/10.5281/zenodo.8202331}}
 ```
+
+Data paper: Schneider, Stefanie (2024). "ARTigo: Data from Social Tagging with Art-historical Images", *Journal of Open Humanities Data* 10, [10.5334/johd.247](https://doi.org/10.5334/johd.247). Code: [arthist-lmu/artigo](https://github.com/arthist-lmu/artigo).