Sign inGet started
Dataset · Hugging Face

EmergentMethods/AskNews-NER-v0

Dataset Card for Dataset Name This dataset aims to improve the representation of underrepresented topics and entities in entity extractors, thereby improving entity extraction accuracy and generalization, especially on the latest news events (dataset represents broad news coverage between February 20-March 31, 2024). The dataset is a collection of news article summaries, translated and summarized with Llama2, and then entities extracted with Llama3. The distribution of data… See the full description on the dataset page: https://huggingface.co/datasets/EmergentMethods/AskNews-NER-v0.

Author / organizationEmergentMethods
License (as declared)apache-2.0
Downloads165 (as reported by Hugging Face)
Likes11
Created19 May 2024
Last updated at source22 May 2024
Embed this entry
<iframe src="https://silytics.si/embed/item/8e708a96-67c4-4938-a94e-e407e3e0ede0" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 2 days

165 now (-1 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

Connected in the knowledge graph

Explore this entry in the graph

Recent changes

  • 16 Oct 2026 Description: Dataset Card for Dataset Name This dataset aims to improve the representation of underrepresented topics and entities in entity extractors, thereby improving entity extraction accuracy and generalization, especially on the latest news event → Dataset Card for Dataset Name This dataset aims to improve the representation of underrepresented topics and entities in entity extractors, thereby improving entity extraction accuracy and generalization, especially on the latest news event

Changes detected by Silytics between refreshes of the source.

More from EmergentMethods

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.