Datasets:

V4ldeLund
/

da-wiki-icc

url string	image_urls string	images unknown	captions string	neighbouring_context string	row_id int64	full_text_row_id int64	has_full_text bool	full text string
https://da.wikipedia.org/wiki/%22Krusche_i_Ender%22_murstenshus	https://upload.wikimedia…i_Ender_Lodz.jpg	"/9j/4AAQSkZJRgABAQEAYABgAAD/4TcdRXhpZgAASUkqAAgAAAAUAA8BAgAKAAAA/gAAABABAgAJAAAACAEAABoBBQABAAAAEgE(...TRUNCATED)	"Krusche i Ender" murstenshus	"[ ] 2 sprog * [Norsk bokmål](https://no.wikipedia.org/wiki/%C2%ABKrusche_i_Ender%C2%BB_mursteinshu(...TRUNCATED)	0	0	true	"# \"Krusche i Ender\" murstenshus\n\n[ ]\n\n2 sprog\n\n* [Norsk bokmål](https://no.wikipedia.org/w(...TRUNCATED)
https://da.wikipedia.org/wiki/%22Superstar%22_Billy_Graham	https://upload.wikimedia…-_March_1975.jpg	"/9j/4AAQSkZJRgABAQAASABIAAD//gAyUHJvY2Vzc2VkIEJ5IGVCYXkgd2l0aCBJbWFnZU1hZ2ljaywgejEuMS4wLiB8fEIy/9s(...TRUNCATED)	Superstar Billy Graham i storhedsperioden, 1975.	"[ ] 22 sprog * [العربية](https://ar.wikipedia.org/wiki/%D8%A8%D9%8A%D9%84%D9%8A_%D8%BA%D8%B1(...TRUNCATED)	1	1	true	"# \"Superstar\" Billy Graham\n\n[ ]\n\n22 sprog\n\n* [العربية](https://ar.wikipedia.org/wiki(...TRUNCATED)
https://da.wikipedia.org/wiki/%22Weird_Al%22_Yankovic	https://upload.wikimedia…rdAlYankovic.jpg	"/9j/4AAQSkZJRgABAQAAAQABAAD/2wBDAAsICAoIBwsKCQoNDAsNERwSEQ8PESIZGhQcKSQrKigkJyctMkA3LTA9MCcnOEw5PUN(...TRUNCATED)	"Weird Al" Yankovic	"## Diskografi\n\n[[redigér](/w/index.php?title=%22Weird_Al%22_Yankovic&veaction=edit&section=2 \"R(...TRUNCATED)	2	2	true	"# \"Weird Al\" Yankovic\n\n[ ]\n\n83 sprog\n\n* [Afrikaans](https://af.wikipedia.org/wiki/%22Weird_(...TRUNCATED)
https://da.wikipedia.org/wiki/%27Allo_%27Allo%21	https://upload.wikimedia…2011_bild_15.jpg	"/9j/4AAQSkZJRgABAQAAAQABAAD/4RiQRXhpZgAATU0AKgAAAAgADAEOAAIAAAAOAAAAngEPAAIAAAAHAAAArAEQAAIAAAALAAA(...TRUNCATED)	Skuespilleren Richard Gibson som gestapos Herr Flick	"## Medvirkende\n\n[[redigér](/w/index.php?title=%27Allo_%27Allo!&veaction=edit&section=1 \"Redigé(...TRUNCATED)	3	3	true	"# 'Allo 'Allo!\n\n[ ]\n\n29 sprog\n\n* [Беларуская](https://be.wikipedia.org/wiki/%D0%9(...TRUNCATED)
https://da.wikipedia.org/wiki/%27Ndrangheta	https://upload.wikimedia…ta_Europe-v4.png	"iVBORw0KGgoAAAANSUhEUgAACDQAAAX1CAIAAAB4VY9PAAAABGdBTUEAALGPC/xhBQAACjppQ0NQUGhvdG9zaG9wIElDQyBwcm9(...TRUNCATED)	'Ndrangehtas udbredelse i Europa	"[ ] 42 sprog * [Afrikaans](https://af.wikipedia.org/wiki/%27Ndrangheta \"'Ndrangheta – afrikaans\(...TRUNCATED)	4	4	true	"# 'Ndrangheta\n\n[ ]\n\n42 sprog\n\n* [Afrikaans](https://af.wikipedia.org/wiki/%27Ndrangheta \"'Nd(...TRUNCATED)
https://da.wikipedia.org/wiki/%27s-Hertogenbosch	https://upload.wikimedia…raat_2A_tm_C.JPG	"/9j/4AAQSkZJRgABAQEASABIAAD/4VMCRXhpZgAATU0AKgAAAAgADwEOAAIAAAAgAAAIzgEPAAIAAAAFAAAI7gEQAAIAAAAJAAA(...TRUNCATED)	Hus i 's-Hertogenbosch	"## Galleri\n\n[[redigér](/w/index.php?title=%27s-Hertogenbosch&veaction=edit&section=1 \"Redigér (...TRUNCATED)	5	5	true	"# 's-Hertogenbosch\n\n[ ]\n\n92 sprog\n\n* [Afrikaans](https://af.wikipedia.org/wiki/%27s-Hertogenb(...TRUNCATED)
https://da.wikipedia.org/wiki/%27s-Hertogenbosch	https://upload.wikimedia…andenboschnl.jpg	"/9j/4X/8RXhpZgAASUkqAAgAAAAMAA4BAgANAAAAngAAAA8BAgAQAAAArAAAABABAgAQAAAAvAAAABIBAwABAAAAAQAAABoBBQA(...TRUNCATED)	Kathedralen i 's-Hertogenbosch	"## Galleri\n\n[[redigér](/w/index.php?title=%27s-Hertogenbosch&veaction=edit&section=1 \"Redigér (...TRUNCATED)	6	5	false	null
https://da.wikipedia.org/wiki/%27s-Hertogenbosch	https://upload.wikimedia…sch-OpenTopo.jpg	"/9j/4AAQSkZJRgABAQEAYABgAAD//gAKQXV0aG9yOiD//gALQ3JlYXRlZDog//4AHUNyZWF0b3I6IFFHSVMgMy4yNC4wLVRpc2x(...TRUNCATED)	Topografisk kommunekort december 2015	"## Galleri\n\n[[redigér](/w/index.php?title=%27s-Hertogenbosch&veaction=edit&section=1 \"Redigér (...TRUNCATED)	7	5	false	null
https://da.wikipedia.org/wiki/%27s-Hertogenbosch	https://upload.wikimedia…emeente_1866.jpg	"/9j/4AAQSkZJRgABAQEBLAEsAAD/2wBDAAUDBAQEAwUEBAQFBQUGBwwIBwcHBw8LCwkMEQ8SEhEPERETFhwXExQaFRERGCEYGh0(...TRUNCATED)	's-Hertogenbosch i 1866	"## Galleri\n\n[[redigér](/w/index.php?title=%27s-Hertogenbosch&veaction=edit&section=1 \"Redigér (...TRUNCATED)	8	5	false	null
https://da.wikipedia.org/wiki/(136617)_1994_CC	https://upload.wikimedia…moons-static.jpg	"/9j/4AAQSkZJRgABAgEBLAEsAAD/4Q3VRXhpZgAASUkqAAgAAAAPAAABAwABAAAAQgMAAAEBAwABAAAArQEAAAIBAwAEAAAAwgA(...TRUNCATED)	Billeder af (136617) 1994 CC med de to satellitter i kredsløb om asteroiden	"[ ] 13 sprog * [العربية](https://ar.wikipedia.org/wiki/%28136617%29_1994_CC \"(136617) 1994 (...TRUNCATED)	9	9	true	"# (136617) 1994 CC\n\n[ ]\n\n13 sprog\n\n* [العربية](https://ar.wikipedia.org/wiki/%28136617(...TRUNCATED)

End of preview. Expand in Data Studio

Dataset Card for “Danish Wikipedia — Image, Caption, Context”

Summary

This dataset contains images from Danish Wikipedia articles paired with:

captions — the local image caption
neighbouring_context — the surrounding section/block text where the image appears in markdown
full text — the full article markdown, stored once per article, with all other rows pointing to the canonical copy via full_text_row_id

Dataset Structure

Files & Splits

Single split: train
Files: data/merged_*.parquet (~5 GiB each)

Important: The pointer columns row_id and full_text_row_id are valid within each merged file.

Fields

Field	Type	Description
`url`	string	Wikipedia article URL.
`image_urls`	string	Image URL
`images`	binary	Raw image bytes; cast to `datasets.Image()` at load time.
`captions`	string	Cleaned image caption text.
`neighbouring_context`	string	Section text near the image.
`row_id`	int64	Row index within the merged file.
`full_text_row_id`	int64	Row id within the same merged file that holds the article’s full text.
`has_full_text`	bool	`True` only for the first row per article.
`full text`	string	Full article markdown, present only where `has_full_text == True`.

Dataset Creation

Rationale

Previous Image-text datasets like alexandrainst/nordjylland-news-image-captioning and alexandrainst/da-withave only captions as context for every image. With recent advancements in Vision Language modelling, we think it is important to have more context to improve the performance of the models. We are hoping that the dataset will become a foundation for Danish VLMs and Vision-language benchmarks.

Source Data

Articles and Images: Danish Wikipedia

Data Curation

The dataset is built from scraping Wikipedia and doing HTML to Markdown conversion, and then images and captions were extracted using regular expressions. For markdown conversion MarkItDown was used

Pipeline Overview

1 Fetch & Convert

Download the article HTML and convert it to Markdown using MarkItDown.

Markdown Cleaning
- Trim to the first # heading (article title) onward.
- Drop typical ending sections: Noter/Kilder/Referencer/Litteratur/Ekstern(e)/External links/References/Notes/Se også/Further reading.
- Remove inline citation footnotes (e.g., #cite_note, #cite_ref) and normalize whitespace.
Image Extraction
- Parse Markdown lines for:
  - Plain images (![alt](url "title"))
  - Linked images ([![alt](url "title")](href "title"))
  - Gallery bullets (- ![...](...))
Neighbouring Context
- Identify H2 section at the top of the image after markdown parsing and take this section as neighboring context
- If the image appears before any H2 sections, take the first few intro sentences.
Optimizing
- Only the first row per article retains the full article markdown.
- All other rows for that article set "full text" = False and point to the row containing full text via: full_text_row_id.
Splitting to ~5 GiB Parts
- For efficiency, the dataset was divided into 78 .parquet files (each ~5 GiB )

Comparison to existing sources

Quantitative comparison

alexandrainst/nordjylland-news-image-captioning: 11.7k image pairs
alexandrainst/da-wit: 168k image pairs
V4ldeLund/da-wiki-icc: 170k image pairs

Qualitative comparison

By focusing on one language, we have managed to create a more accurate parsing pipeline for the Danish Wikipedia. For example, in the Wikipedia article for Aarhus Hovedbanegård our dataset contains 29 images as opposed to only 9 in da-wit.