Wikinews Zhtw: Traditional Chinese News Articles from Wikimedia Dumps

Traditional Chinese Wikinews articles, a collaborative journalism project operated by volunteers under the Wikimedia Foundation. The dataset is constructed from Wikimedia dumps of the 'zhwikinews' project, with each example containing a cleaned, full article. The dataset card was last updated on 2023-10-10.

Use Cases

• Training language models for Traditional Chinese text generation based on news-style writing.
• Analyzing collaborative journalism patterns and article structure within the Wikinews platform.
• Conducting sentiment or topic analysis on a corpus of volunteer-written news articles.
• Benchmarking text cleaning or information extraction pipelines on semi-structured wiki markup.

Strengths

• Data is sourced from the authoritative Wikimedia dumps, ensuring provenance from a major open-knowledge project.
• Articles are described as cleaned to remove unwanted parts, suggesting some preprocessing for usability.
• The dataset focuses on a specific language and project (Traditional Chinese Wikinews), providing a defined scope.

Limitations

• Row count, column definitions, file formats, and license information are all unknown, hindering suitability assessment.
• The description metadata is limited; actual data structure and quality require manual inspection after download.
• Last updated 2023-10-10 04:06:53; freshness should be verified against more recent Wikimedia dumps.

Provenance

Source

Wikimedia dumps (https://dumps.wikimedia.org/zhwikinews)

Collection Method

Constructed from periodic website data dumps, with cleaning applied to remove unwanted parts.

Time Range

Temporal coverage of the articles is not specified, but dumps are periodic.

Freshness

Dataset card last updated 2023-10-10 04:06:53.

Geography

Not explicitly stated, but content is from the Traditional Chinese Wikinews project.