jvdgoltz/dbnl.org-dutch-public-domain

Dataset Card for "dbnl.org-dutch-public-domain"

Table of Contents

• —Dataset Description
• —Dataset Summary
• —Supported Tasks and Leaderboards
• —Languages
• —Dataset Structure
• —Data Instances
• —Data Fields
• —Data Splits
• —Dataset Creation
• —Curation Rationale
• —Source Data
• —Annotations
• —Personal and Sensitive Information
• —Considerations for Using the Data
• —Social Impact of Dataset
• —Discussion of Biases
• —Other Known Limitations
• —Additional Information
• —Dataset Curators
• —Licensing Information
• —Citation Information
• —Contributions

Dataset Description

• —Homepage: DBNL Public Domain Collection
• —Point of Contact: julian at vdgoltz.net

Dataset Summary

This dataset comprises a collection of texts from the Dutch Literature in the public domain, specifically from the DBNL (Digitale Bibliotheek voor de Nederlandse Letteren) public domain collection. The collection includes books, poems, songs, and other documentation, letters, etc., that are at least 140 years old and thus free of copyright restrictions. Each entry in the dataset corresponds to one section of a chapter of a text, ensuring a granular level of detail for text analysis.

Supported Tasks and Leaderboards

• —Language Modeling
• —Text Generation
• —Other tasks that can benefit from historical Dutch texts

Languages

The dataset is primarily in Dutch (nl).

Dataset Structure

Data Instances

A data instance corresponds to a section of a chapter of a document, including metadata such as title, author, publication year, and the text content itself.

Data Fields

• —ti_id: Unique text identifier
• —titel: Title of the text
• —jaar: Publication year
• —druk: Edition
• —bibliotheek: Library code
• —categorie: Category ID
• —pers_id: Person ID
• —voornaam: Author's first name
• —achternaam: Author's last name
• —url: URL to the text
• —text_url: URL to the text in .txt format
• —revision_date: Date of the revision
• —edition: Edition details
• —language: Language of the text
• —chapter: Chapter number
• —section: Section number

Data Splits

The dataset is split into training and validation sets at text level (90:10), ensuring that sections or chapters from the same document do not leak from one split to another.

Dataset Creation

Curation Rationale

The dataset was curated to make historical Dutch texts available for computational analysis, preserving cultural heritage and supporting research in the humanities and linguistic studies.

Source Data

Initial Data Collection and Normalization

Data was collected from the DBNL's public domain collection, normalized, and structured to facilitate computational use.

Who are the source language producers?

The source language producers are authors of Dutch literature whose works have entered the public domain, implying their passing at least 70 years ago.

Annotations

The dataset does not contain annotations.

Personal and Sensitive Information

Given the historical nature of the texts, they are free from personal and sensitive information concerns in the contemporary sense. However, they reflect the social norms, biases, and cultural contexts of their time.

Considerations for Using the Data

Social Impact of Dataset

The dataset serves as a valuable resource for understanding Dutch literary heritage, cultural history, and language evolution over time. It can support diverse research agendas in computational linguistics, cultural studies, and history.

Discussion of Biases

The texts contain biases prevalent at their time of publication, including colonialism, racism, sexism, and other societal norms of their era. Users are urged to consider these contexts critically and use the data responsibly.

Other Known Limitations

The dataset's historical nature means it may not be suitable for applications requiring contemporary language use or norms.

Additional Information

Dataset Curators

This dataset was curated by https://huggingface.co/jvdgoltz, who is not affiliated with DBNL.org and did not act on their behalf. The data is sourced from the DBNL public domain collection.

Licensing Information

The texts in this dataset are in the public domain. According to Chat-GPT 4, the best fitting license would be: Creative Commons Zero v1.0 Universal, making them legally available for use, redistribution, and adaptation by anyone for any purpose.

Citation Information

Not applicable.