SinclairSchneider/focus_de

FOCUS online (focus.de)

33 521 German-language news articles crawled from www.focus.de and published as a Hugging Face dataset. This is part of the German Media collection, a set of comparable corpora covering the German media landscape from public-service broadcasters to alternative and fringe outlets, all built with the same crawler and the same extraction pipeline so that they can be compared and combined directly.

⚠️ For scientific research only

This corpus is published exclusively for scientific research under the German research exception for text and data mining — the Forschungsprivileg of § 60d UrhG (implementing Art. 3 of EU Directive 2019/790). The article texts remain the copyright of their publisher and are not licensed for commercial use, redistribution, or republication. By downloading it you confirm that you use it for non-commercial scientific research. See Licensing and intended use.

The articles are reproduced verbatim from the publisher's website; their content expresses the views of that publisher, not of the dataset author.

At a glance

Source	www.focus.de
Rows	33 521
Columns	6 (title, content, category, author, description, date_published_time_at)
Language	German (de)
Split	train (single split)
Format	Parquet, 1 shard, ≈ 96 MB download / 155 MB in memory
Perceived partisanship	4.9 / 7 → right of centre (medienkompass.org reader survey)
Intended use	Scientific research only (Forschungsprivileg, § 60d UrhG)
Mirror	NLP-UniBW/focus_de

from datasets import load_dataset

ds = load_dataset("SinclairSchneider/focus_de", split="train")
print(ds[0]["title"])

About the source

FOCUS online is the news portal of the weekly news magazine FOCUS, which Hubert Burda Media launched in 1993 as a competitor to DER SPIEGEL; the website has been online since 1996. Both belong to the Munich-based Hubert Burda Media group. FOCUS online is one of the most-visited German news sites and is financed by advertising. Besides politics and business it has a strong service and reach-oriented profile (health, finance, motoring, travel, consumer advice, agency and police reports). Editorially the brand is generally regarded as centre-right.

Perceived political position

The German media compass at medienkompass.org places German mainstream and alternative outlets on two axes on the basis of a reader survey (n ≈ 1 154 participants in total, building on a media map by the think tank polisphere; the quality axis is adapted from Vanessa Otero's media chart).

• x axis — Parteilichkeit (partisanship), 1–7: < 3.5 left, 3.5 – 4.5 relatively neutral, > 4.5 right. 1 far-left (fake news and conspiracy theories) · 2 left mission (questionable journalistic standards) · 3 leaning left · 4 minimal partisan tendency · 5 leaning right · 6 right mission (questionable journalistic standards) · 7 far-right (fake news and conspiracy theories).
• y axis — Qualität (quality), 1–5: 1 sensationalist / clickbait · 2 basic information · 3 meets high standards · 4 analytical · 5 complex.

In that survey, Focus Online is rated:

	Score	Reading
Partisanship (1–7)	4.9	right of centre
Quality (1–5)	1.7	on the 1 = clickbait … 5 = complex scale

Parteilichkeit (partisanship):  1 = far left  ...  7 = far right

1         2         3         4         5         6         7
|---------|---------|---------|---------|---------|---------|
                                       ^                     
                                 Focus Online = 4.9

left  <--------------|  neutral  |-------------->  right
                   3.5           4.5

For context, here is the full survey table, sorted from left to right — the other outlets in the collection are directly comparable:

Outlet	Partisanship (1–7)	Quality (1–5)
indymedia	2.0	1.7
Jungle World	2.3	2.3
junge Welt	2.4	2.3
neues Deutschland	2.6	2.3
der Freitag	2.7	2.9
die Tageszeitung	2.8	2.7
Vice	2.8	1.8
NachDenkSeiten	3.1	3.0
Frankfurter Rundschau	3.4	2.8
SPON	3.5	2.1
Spiegel Print	3.5	2.9
Süddeutsche	3.5	3.3
WDR	3.5	2.8
Tagesspiegel	3.6	2.8
Zeit	3.6	3.5
ARD	3.7	2.9
NDR	3.7	2.8
Deutschlandfunk	3.8	3.5
Stern	3.8	2.1
ZDF	3.8	2.8
MDR	4.1	2.5
Handelsblatt	4.3	3.3
NTV	4.3	2.1
BR	4.4	2.6
N24	4.4	2.0
Wirtschaftswoche	4.4	3.1
Frankfurter Allgemeine	4.5	3.3
RTL	4.5	1.5
Die Welt	4.8	2.5
Cicero	4.9	3.1
Focus Online ⬅ this dataset	4.9	1.7
Deutsche Wirtschafts Nachrichten	5.0	2.2
MM News	5.1	1.7
RT Deutsch	5.1	2.0
Achgut	5.2	2.4
BILD	5.2	1.3
Tichys	5.5	2.6
junge Freiheit	5.8	1.9
Compact	6.0	1.8

How to read these numbers. They are perception data from a self-selected online survey, not a content analysis: participants rated outlets they know, the sample is not representative of the German population, and the site itself describes the project as an early iteration. Treat them as a rough, crowd-sourced orientation — useful for grouping datasets into coarse political blocks, not as ground truth for bias labelling.

Dataset structure

Single train split with 33 521 rows and the following columns:

Column	Type	Description	Extracted from
title	string	Headline of the article as given in the HTML <title> (with - FOCUS online suffix).	the HTML <title> element, including the site suffix - FOCUS online
content	string	Full plain-text body of the article.	get_text() of <div class="articleContent"> after removing div.article-social-border and div.affiliate_info; tabs removed, runs of 3+ newlines collapsed to two, repeated spaces to one
category	string	Top-level section derived from the canonical URL; empty when not one of the 17 known sections.	first path segment of the <link rel="canonical"> URL if it is one of 17 known sections, otherwise empty
author	string	Value of <meta name="author">; FOCUS online for every row.	<meta name="author">
description	string	Teaser / meta description of the article.	<meta name="description">
date_published_time_at	timestamp[us, tz=+02:00]	Publication timestamp from <meta name="date">.	<meta name="date">, parsed with dateutil during consolidation

Example record

{
  "title": "Black Friday Wochenende 2022: Diese Top-Deals gibt es nur noch für kurze Zeit",
  "content": "PantherMedia / Sorapop Udomsri\n\nMontag, 28.11.2022, 12:38\n\nDer Cyber Monday steht dem Black Friday in nichts nach: Das Schnäppchen-Karussell dreht sich weiter. Wir haben die besten Deals für Sie zusammengetragen, die sich am letzten Tag lohnen. Das ein oder andere Weihnachtsgeschenk ist mit Sicherheit dabei!\n\nCyber Monday – diese Deals lohnen sich nach dem Black Friday besonders\nIn diesem Artikel  …",
  "category": "kaufberatung",
  "author": "FOCUS online",
  "description": "Nach dem Black Friday ist vor dem Cyber Monday - und das Schnäppchen-Karussell dreht sich weiter. Wir haben die besten Deals für Sie zusammengetragen.",
  "date_published_time_at": "2022-11-28 13:38:29+02:00"
}

Temporal coverage

The date_published_time_at column could be parsed for 33 521 of 33 521 rows (100.0 %). Those rows span 2001-03-28 → 2023-09-27.

Year	Articles
2001	2	█
2004	3	█
2005	5	█
2006	1	█
2007	4	█
2008	13	█
2009	12	█
2010	17	█
2011	31	█
2012	15	█
2013	7 306	████████████████████████
2014	1 661	█████
2015	2 716	█████████
2016	1 401	█████
2017	1 203	████
2018	1 201	████
2019	1 558	█████
2020	1 950	██████
2021	5 231	█████████████████
2022	4 510	███████████████
2023	4 681	███████████████

How the data was collected

The whole pipeline is contained in this repository under scraping/ and consists of two steps: a distributed crawler that walks the site and writes one JSON file per article, and a consolidation notebook that merges those files into the dataset you are reading.

1. Crawling — scraping/crawler.py

The crawler is a breadth-first whole-site crawler built on MechanicalSoup and multiprocessing:

1. Sharded frontier. 32 worker processes each own one multiprocessing.Manager().Queue(). A URL is assigned to a worker by int(md5(url).hexdigest(), 16) % 32, so every URL is always handled by the same worker and no locking or cross-process deduplication is needed.
2. Seed. The crawl is seeded with a single article URL (https://www.focus.de/kaufberatung/deals_angebote/ice-guenstiger-als-regionalbahn-die-bahn-lockt-mit-ice-tickets-fuer-9-90-euro-das-muessen-sie-beachten_id_178649661.html) and discovers everything else by following links.
3. Link discovery. Every fetched page is parsed with BeautifulSoup/lxml; all <a href> values that start with https://www.focus.de/ or with / (which are then made absolute) are added to the frontier. URLs containing # are dropped.
4. URL filtering. Before fetching, a URL is skipped when it contains /files/, /video/, .jpg, .pdf or .epub.
5. Fetching. Each URL gets up to 1000 attempts with a 10 s timeout.
6. Extraction. get_content() turns the HTML into a record (see below). A page counts as an article when it contains a <div class="displayDate"> element; only those are written out, as pages/<flattened-url>.json (the body itself is not required to be non-empty).
7. Resumability. Each worker appends every known and every visited URL to data/thread_<id>_pages.txt and data/thread_<id>_pages_visited.txt, and re-reads them on start, so an interrupted crawl continues where it stopped. Creating a file named killall.txt in the working directory stops all workers cleanly.

Crawler settings used for this site:

• Worker processes: 32 (iNumberOfThreads = 32)
• Retries per URL: 1000, with a 10-second socket timeout per attempt
• Proxy pool: not used — the original crawl connected directly (the published script can still route through proxies via SCRAPER_PROXIES, see below)
• User-agent rotation: no, MechanicalSoup's default UA
• Politeness delay: none — the crawler relied on its retry loop
• Bot-challenge handling: none
• Raw HTML dump: disabled

Paywall handling. The crawler contains no paywall check: every page that passed the article test (see below) was stored, so if a page was paywalled at crawl time only its freely served part can be in the dataset.

2. Article extraction

get_content() reads every field from the page's HTML (no JSON-LD is used):

Field	Source in the page
title	the HTML <title> element, including the site suffix - FOCUS online
content	get_text() of <div class="articleContent"> after removing div.article-social-border and div.affiliate_info; tabs removed, runs of 3+ newlines collapsed to two, repeated spaces to one
category	first path segment of the <link rel="canonical"> URL if it is one of 17 known sections, otherwise empty
author	<meta name="author">
description	<meta name="description">
date_published_time_at	<meta name="date">, parsed with dateutil during consolidation

The 17 sections recognised for category are politik, finanzen, perspektiven, earth, wissen, gesundheit, kultur, panorama, sport, reisen, auto, immobilien, kaufberatung, experts, deals, digital and familie. The crawled URL itself is not stored.

Each article is stored as a JSON file of the shape

{
  "title": "…",
  "content": "…",
  "additional_attributes": { "is_article": true, "category": "…", "author": "…", "description": "…", "date_published_time_at": "…" }
}

3. Consolidation — scraping/finalize.ipynb

1. glob("pages/*.json") collects every article file.
2. title, content and the fields under additional_attributes are flattened into one pandas.DataFrame column each; date_published_time_at is parsed with dateutil.parser.parse.
3. No de-duplication or filtering is done.
4. The frame is written to focus_de.json and pushed with datasets.Dataset.from_pandas(df).push_to_hub("SinclairSchneider/focus_de").

Dataset-specific notes

• title carries the site suffix - FOCUS online in 99.7 % of rows; strip it with str.removesuffix(" - FOCUS online").
• author is the constant FOCUS online for every row: the <meta name="author"> tag does not name the journalist, so the column carries no information.
• content often starts with the photo credit of the lead image (e.g. picture alliance/dpa), because the credit sits inside div.articleContent.
• category is empty for 2 256 rows (6.7 %), namely articles outside the 17 recognised sections.
• No de-duplication. 135 rows are exact title + content duplicates and 703 titles occur more than once.
• There is no url column.
• All timestamps are stored with one fixed +02:00 offset, also for articles published in winter (CET, +01:00).

Reproducing the crawl

pip install mechanicalsoup beautifulsoup4 lxml python-dateutil fake-useragent \
            pandas datasets huggingface_hub tqdm

mkdir focus_de && cd focus_de
curl -L -O https://huggingface.co/datasets/SinclairSchneider/focus_de/resolve/main/scraping/crawler.py

python crawler.py          # writes pages/*.json; `touch killall.txt` to stop
jupyter notebook finalize.ipynb

Routing the crawl through your own proxies

The original crawl connected directly, without proxies. The published crawler.py can nevertheless route its requests through your own proxies: it reads the SCRAPER_PROXIES environment variable: a comma-separated list of proxies, written with or without a scheme (http:// is assumed when none is given). One entry is picked per request, so a pool spreads the load across several exits.

# a single proxy
export SCRAPER_PROXIES="proxy-a.example.com:3128"

# a pool, rotated per request
export SCRAPER_PROXIES="proxy-a.example.com:3128,proxy-b.example.com:3128"

# with an explicit scheme and credentials
export SCRAPER_PROXIES="http://user:pass@proxy.example.com:3128"

# SOCKS, if PySocks is installed
export SCRAPER_PROXIES="socks5://localhost:9050"

# back to a direct connection — this is the default
unset SCRAPER_PROXIES

python crawler.py

Internally this becomes requests' standard {'http': …, 'https': …} proxy mapping; with the variable unset the mapping is empty and the crawler connects directly. No credentials, cookies or personal data are contained in the published code.

Please crawl responsibly: respect robots.txt and the site's terms of use, keep the concurrency far below the 32 workers used here unless you have the site's permission, and prefer this published dataset over re-crawling.

Usage

from datasets import load_dataset
import pandas as pd

ds = load_dataset("SinclairSchneider/focus_de", split="train")

# stream instead of downloading everything
stream = load_dataset("SinclairSchneider/focus_de", split="train", streaming=True)
print(next(iter(stream))["title"])

# as a DataFrame
df = ds.to_pandas()
df["date_published_time_at"] = pd.to_datetime(df["date_published_time_at"], errors="coerce", utc=True, format="mixed")
recent = df[df["date_published_time_at"].dt.year >= 2024]
print(len(recent), "articles from 2024 onwards")

Combining several outlets into one comparative corpus:

from datasets import load_dataset, concatenate_datasets

outlets = ["SinclairSchneider/focus_de", "SinclairSchneider/tagesschau_de"]
parts = []
for name in outlets:
    d = load_dataset(name, split="train").select_columns(["title", "content"])
    parts.append(d.add_column("outlet", [name.split("/")[-1]] * len(d)))
corpus = concatenate_datasets(parts)

Limitations and biases

• Crawl, not archive. Coverage is whatever was reachable by following links from the seed URL at crawl time. Articles that were deleted, depublished or orphaned before the crawl are missing, and old content is covered less densely than recent content.
• No paywall filter. The crawler stored every article page regardless of access status; paywalled pieces, if any, are represented only by their freely served text.
• No de-duplication. See the dataset-specific notes above; deduplicate on title + content before training or counting.
• Noisy metadata. author is constant, category is derived from the URL rather than from the publisher's own section data, and the date offset is not season-aware.
• Editorial perspective. This is the output of a single outlet. Any model trained on it alone inherits that outlet's selection of topics, framing and vocabulary — which is exactly why the collection provides the same pipeline across the political spectrum.
• Light text normalisation only. Whitespace is collapsed, but image credits, captions and occasional residual boilerplate remain in content.

Licensing and intended use

Scientific research only — Forschungsprivileg (§ 60d UrhG)

This dataset is made available solely for the purpose of scientific research. The legal basis is the German text-and-data-mining exception for research, the so-called Forschungsprivileg:

• § 60d UrhG — Text und Data Mining für Zwecke der wissenschaftlichen Forschung — allows research organisations to make reproductions of lawfully accessible works in order to carry out text and data mining for non-commercial scientific research, and to share the resulting corpus with a specifically limited circle of people for joint research and with individual third parties for the purpose of quality assurance.
• It implements Art. 3 of Directive (EU) 2019/790 (the DSM Directive); comparable research exceptions exist in the other EU member states.
• § 60d UrhG does not cover commercial exploitation. Research organisations that give a commercial enterprise preferential access are explicitly excluded from the privilege.

What this means for you

✅ Permitted — using the corpus for non-commercial scientific research: text and data mining, corpus linguistics, media-bias and discourse analysis, training and evaluating research models, reproducing published findings, and quality checks by individual third parties.

❌ Not permitted — commercial use or commercial model training; redistributing or republishing the article texts; presenting the texts as your own; building products or services on the article texts; any use that substitutes for reading the publisher's own website.

The article texts remain the copyright of FOCUS online and its publisher. No licence to them is granted or implied by this repository. Only the crawling and extraction code in scraping/ is offered for free reuse.

If your intended use falls outside the research exception, obtain permission from the rights holder, or use only the derived, non-expressive data (statistics, embeddings, counts). Note that § 60d UrhG also expects research corpora to be deleted once they are no longer needed for the research purpose, with archiving permitted at trusted institutions.

If you are a rights holder and want content removed, please open a discussion on this repository.

This section is an orientation for researchers, not legal advice. Responsibility for your own use of the data remains with you and your institution.

Citation

@misc{focus_de_dataset,
  title        = {FOCUS online (focus.de): a corpus of German news articles},
  author       = {Schneider, Sinclair},
  year         = {2023},
  howpublished = {Hugging Face},
  url          = {https://huggingface.co/datasets/SinclairSchneider/focus_de}
}

See also

• The full German Media collection
• Mirror of this dataset: NLP-UniBW/focus_de
• medienkompass.org — Deutsche Medienlandschaft, the source of the perceived-partisanship figures above