Datasets:
Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
Hebrew
Size:
10K - 100K
Tags:
legal
metadata
task_categories:
- text-generation
language:
- he
tags:
- legal
pretty_name: Knesset News Dataset
size_categories:
- 10K<n<100K
Dataset Card for Knesset News
Hebrew Press Releases From the Knesset(Israeli Parliament)
Dataset Details
Dataset Description
The dataset contains all the official knesset hebrew press releases up until 21-07-2025
- Curated by: [@Wissotsky]
- Language: [Hebrew]
Dataset Sources
Dataset Structure
- SP_Id (string): Unique identifier for each news article from the Knesset system
- Title (string): Article headline/subject
- Date (string): Publication date of the article
- URL (string): Original source URL of the article on the Knesset website. (relative link)
- Content (string): Full article content converted to Markdown format, including extracted title, date, and main article text from the raw html data.
Dataset Creation
Data Collection and Processing
The title,date and article text were extracted from the html data, concatenated and converted to markdown using the html-to-markdown/v2 golang library
Bias, Risks, and Limitations
Well its the Knesset news, whaddya expect