Dataset · Harrysunshine
OpenAlex abstracts dataset (1980–2026)
Harrysunshine/openalex-abstracts-106m
Massive corpus of over 106 million English scientific articles and conference papers.
What's inside
Contains over 106 million rows with abstracts of scientific articles and conference papers from 1980 to 2026. Data is presented in tabular and text format in English, gathered from OpenAlex, OpenAIRE, Semantic Scholar, and PubMed.
What you can do with it
- Training language models on scientific texts
- Text classification and feature-extraction tasks
- Tasks
- Text classification, Embeddings
- Languages
- English
- Size
- 100M–1B rows
- Downloads this month
- 21.7K
- Likes
- 1
- Access
- on request on Hugging Face
How to load it
from datasets import load_dataset
ds = load_dataset("Harrysunshine/openalex-abstracts-106m")