Skip to content

Dataset · TimSchopf

arXiv article titles and abstracts across 130 scientific categories

TimSchopf/arxiv_categories

Over 200k arXiv paper titles and abstracts labeled across 130 thematic classes.

What's inside

The dataset is based on arXiv scientific library metadata and contains 203,961 English title-abstract pairs. Each article is assigned labels from 130 different scientific categories for multi-label classification. The set was prepared by the authors of a scientific paper for the ICNLSP 2024 conference.

What you can do with it

  • Training multi-label classification models for scientific and technical texts with a large number of classes
  • Testing few-shot learning algorithms on article topic labeling
Tasks
Text classification
Languages
English
Size
100K–1M rows
Downloads this month
910
Likes
6

How to load it

from datasets import load_dataset

ds = load_dataset("TimSchopf/arxiv_categories")