ArXiv Paper Metadata Spanning 1991 to 2026

Open arXiv contains metadata for 2.99 million scientific papers from the arXiv preprint repository. The collection includes titles, abstracts, authors, categories, DOIs, and version history, covering publications from 1991 to 2026.

Use Cases

• Search and filter papers by author, category, or title for literature review and discovery.
• Analyze version history and publication trends across 40 years of scientific output.
• Link papers to external resources using provided DOIs for citation network analysis.
• Train text classification models on paper abstracts and assigned category labels.

Strengths

• Contains 2.99 million paper records, providing a substantial corpus for analysis.
• Covers a 35-year time range from 1991 to 2026, enabling longitudinal study.
• Includes diverse metadata fields such as abstracts, authors, and categories for multifaceted querying.

Limitations

• The specific column structure and data types are not documented in the input.
• Sample data is unavailable, making initial data quality assessment difficult prior to download.

Provenance

Source

Cornell University arXiv repository.

Collection Method

Converted from the Cornell University Kaggle dataset into Parquet format.

Time Range

1991 to 2026

Freshness

Dataset was last updated on 2026-03-24.