Scholarly sources¶
Four connectors read your papers and citations from public scholarly databases: Semantic Scholar, OpenAlex, arXiv and ORCID.
None of them needs a key or an account. Add one or several; the same paper found by more than one is proposed only once.
How they work¶
Each connector turns an author profile into:
- One author item, with profile metrics where the source has them:
citations,h_indexandpapers. - One paper item per work (up to 500), with its
citationswhen the source counts them.
Every paper is proposed for the scholarly articles criterion, with what the source says about it: venue, year, type, citations, DOI or arXiv id.
| What the source says | Proposed as | Stage |
|---|---|---|
| A journal | Journal article (peer reviewed) | Published |
| A conference | Conference paper (peer reviewed) | Published |
| A preprint, or no venue | Preprint | Preprint |
| A named venue of unknown type | Paper | Published |
A preprint's proposal asks you to add the venue if it was later published.
Confirm each paper is yours
Author pages get merged, and people share names. A connector can't know whether a listed paper is really yours, so every proposal asks you to confirm you're an author before you accept it.
Papers are matched across sources by arXiv id, then DOI. A paper linked from a GitHub README or a Hugging Face repo is the same candidate as the one a scholarly connector finds.
Semantic Scholar¶
Uses the free Semantic Scholar Graph API. Reads your papers, citations, venues, publication types and h-index.
| Input | Example |
|---|---|
| Author page | https://www.semanticscholar.org/author/Maya-Chen/12345678 |
| Handle | s2:12345678 |
Find your author page by searching your name on semanticscholar.org. Requests are rate-limited; Area O1 backs off and retries.
OpenAlex¶
Uses the free OpenAlex API. Reads your works, citation counts, venues, types and h-index.
| Input | Example |
|---|---|
| Author URL | https://openalex.org/A5000000001 (or https://openalex.org/authors/A5000000001) |
| Handle | openalex:A5000000001 |
OpenAlex asks callers for a contact email for its faster "polite pool". Area O1 doesn't send one, because it would identify you.
arXiv¶
Reads the papers on your arXiv author page, from its Atom feed.
| Input | Example |
|---|---|
| Author page | https://arxiv.org/a/chen_m_1 |
| Handle | arxiv:chen_m_1 |
arXiv has no citation counts, so you get the paper list and the papers count. Papers are preprints unless the
feed carries a journal reference.
Because an arXiv author page is claimed by its owner, its proposals start with higher confidence than Semantic Scholar's or OpenAlex's. The same goes for ORCID.
ORCID¶
Reads the works on a public ORCID record through the public ORCID API.
| Input | Example |
|---|---|
| Record URL | https://orcid.org/0000-0002-1825-0097 |
| Bare iD | 0000-0002-1825-0097 or orcid:0000-0002-1825-0097 |
ORCID lists works but not citations. Works are classed by their ORCID type (journal article, conference paper, preprint, working paper). Only works you've made public on your record are visible.
Which ones to add¶
You don't need all four. A common setup:
- ORCID or arXiv for the list of works you've curated yourself.
- Semantic Scholar or OpenAlex for citation counts and h-index over time, on the Metrics page.
Citation numbers differ between databases. Use the same source consistently when you compare over time.