| | --- |
| | license: other |
| | license_name: 20minuten-copyright-notice |
| | license_link: LICENSE |
| | language: |
| | - de |
| | - fr |
| | tags: |
| | - news |
| | - comparable |
| | - comparable corpus |
| | - parallel |
| | - bitext |
| | pretty_name: 20min-XD |
| |
|
| | configs: |
| | - config_name: articles |
| | data_files: "data/articles.tsv" |
| | - config_name: sentences |
| | data_files: "data/sentences.tsv" |
| | --- |
| | |
| | # *20min-XD*: A Comparable Corpus of Swiss News Articles |
| |
|
| | ## Dataset Summary |
| |
|
| | *20min-XD* (**20 Min**uten **cross**-lingual **d**ocument-level) is a comparable corpus of Swiss news articles in German and French, collected from the online editions of [20 Minuten](20min.ch) and [20 minutes](20min.ch/fr) between 2015 and 2024. The dataset consists of 15,000 semantically aligned German and French article pairs. Unlike parallel corpora, *20min-XD* captures a broad spectrum of cross-lingual similarity, ranging from near-translations to related articles covering the same event. This dataset is intended for non-commercial research use only – please refer to the accompanying license/copyright notice for details. |
| |
|
| | ## Details |
| | * Languages: German (de), French (fr) |
| | * Available versions: |
| | * Document level: Contains full article pairs. |
| | * Size: 15,000 article pairs |
| | * Sentence level: Contains alignable sentences from the 15'000 article pairs. Note that all alignable sentences are contained in this dataset without considering a similarity threshold. We recommend, setting a similarity threshold (e.g.: 79) to filter for semantically similar sentences. |
| | * Size: 117,126 sentence pairs |
| | * The datasets contain the following information: |
| | * Document level: |
| | * ```id_de/fr```: unique article ID |
| | * ```score```: cosine similarity score |
| | * ```content_id_de/fr```: content ID |
| | * ```pubtime_de/fr```: time of publication |
| | * ```article_link_de/fr```: link to the online news article |
| | * ```medium_code_de/fr```: abbreviation for the name of the news outlet |
| | * ```medium_name_de/fr```: full name of the news outlet |
| | * ```language_de/fr```: language code |
| | * ```char_count_de/fr```: character count of the content |
| | * ```head_de/fr```: article head text |
| | * ```subhead_de/fr```: article subhead text |
| | * ```content_de/fr```: article content text |
| |
|
| | * Sentence level: |
| | * ```id_de/fr```: unique sentence ID |
| | * ```score```: cosine similarity score |
| | * ```sentence_id_de/fr```: document internal sentence ID |
| | * ```aligned_article_id_de/fr```: document ID |
| | * ```sentence_de/fr```: sentence |
| | * ```char_count_de/fr```: character count of the sentence |
| | * ```char_count_diff```: absolute difference between character counts of German and French sentence |
| |
|
| | # Citation |
| | If you use *20min-XD* in your research, please cite: |
| |
|
| | ``` |
| | @inproceedings{wastl-et-al-2025-20min, |
| | title = "20min-{XD}: A Comparable Corpus of {S}wiss News Articles", |
| | author = "Wastl, Michelle and |
| | Vamvas, Jannis and |
| | Calleri, Selena and |
| | Sennrich, Rico", |
| | editor = {Gerber, Jonathan and |
| | Cieliebak, Mark and |
| | Tuggener, Don and |
| | H{\"u}rlimann, Manuela}, |
| | booktitle = "Proceedings of the 10th edition of the Swiss Text Analytics Conference", |
| | month = may, |
| | year = "2025", |
| | address = "Winterthur, Switzerland", |
| | publisher = "Association for Computational Linguistics", |
| | url = "https://aclanthology.org/2025.swisstext-1.1/", |
| | pages = "1--10" |
| | } |
| | ``` |
| |
|
| |
|
| |
|
| |
|
