[Paper Review] OpenAlex Snapshot

Jason Priem, Heather Piwowar+6 more|arXiv (Cornell University)|May 4, 2022

Topic Modeling6 references271 citations

TL;DR

OpenAlex is a fully open scientific knowledge graph that replaces MAG, providing metadata for millions of works, authors, venues, institutions, and concepts via GUI, data dumps, and REST API.

ABSTRACT

OpenAlex is an open, comprehensive index of scolarly papers, citations, authors, institutions, and journals. Available through API and UI as well (at openalex.org), this record refers to the full data snapshot. When citing OpenAlex, don't use this record. Instead, use: Priem, J., Piwowar, H., & Orr, R. (2022). OpenAlex: A fully-open index of scholarly works, authors, venues, institutions, and concepts. ArXiv. https://arxiv.org/abs/2205.01833 This record is intended for long-term persistence but because the OpenAlex snapshot updates every month, it is better to download the current version directly from AWS. Information on how to download the entire data snapshot for OpenAlex can be found at: https://docs.openalex.org/download-all-data/openalex-snapshot

Motivation & Objective

• Motivate the need for an open, comprehensive scholarly knowledge graph to replace discontinued MAG.
• Describe the structure and scale of OpenAlex, including works, authors, venues, institutions, and concepts.
• Highlight how OpenAlex is accessible and usable via multiple delivery modalities and its ongoing development trajectory.

Proposed method

• Define the OpenAlex knowledge graph as a fully open resource for scholarly metadata.
• Enumerate the components: works, authors, venues, institutions, and Wikidata concepts linked via a classifier.
• Provide access modalities: web GUI, full data dump, and high-volume REST API.
• Discuss current limitations and planned improvements in citation accuracy and entity parsing/deduplication.

Experimental results

Research questions

• RQ1 What are the core components and scale of the OpenAlex knowledge graph (works, authors, venues, institutions, concepts)?
• RQ2 How can researchers access and use OpenAlex (GUI, data dump, API) and what are the current limitations?
• RQ3 What improvements are planned for citation accuracy, author/institution parsing, and deduplication in OpenAlex?

Key findings

• OpenAlex contains metadata for 209 million works (journal articles, books, etc).
• It includes 2013 million disambiguated authors.
• It catalogs 124 thousand venues and 109 thousand institutions.
• It links 65 thousand Wikidata concepts to works via an automated hierarchical multi-tag classifier.