Crossref Scholarly Works — Papers by Query, Journal or DOI avatar

Crossref Scholarly Works — Papers by Query, Journal or DOI

Pricing

from $1.00 / 1,000 results

Go to Apify Store
Crossref Scholarly Works — Papers by Query, Journal or DOI

Crossref Scholarly Works — Papers by Query, Journal or DOI

Search 190M+ scholarly works via Crossref and get flat rows — title, authors, journal, publisher, year, citation count and DOI link.

Pricing

from $1.00 / 1,000 results

Rating

0.0

(0)

Developer

alaudin burki

alaudin burki

Maintained by Community

Actor stats

0

Bookmarked

2

Total users

1

Monthly active users

22 days ago

Last modified

Categories

Features

Limited permissions

Share

Crossref is the registration agency that most academic publishers use to deposit DOI metadata. This Actor searches its works endpoint and writes one flat row per work: title, authors, journal, publisher, publication year, citation count and the DOI link. Filters are passed straight through in Crossref's own syntax, so a search can be narrowed by date, type, journal or publisher without extra inputs.

No key is needed. The Actor pages with Crossref's cursor, 100 works per request, pacing requests 200 ms apart and retrying a throttled or failed request up to three times.

How to use

  1. Enter a free-text Query, a Crossref Filter, or both. Either one alone is enough to search.
  2. Set Max rows and Max requests to bound the run.
  3. Run the Actor and export the dataset; each row links to the work through its DOI.

Inputs

FieldPurposeDefault
queryFree-text bibliographic search over the metadata Crossref holds, including titles, authors and abstracts where the publisher deposited them.—
filterA Crossref filter expression, comma separated, e.g. from-pub-date:2026-01-01,type:journal-article. Passed to Crossref unchanged, so its own filter names and syntax apply.—
maxItemsHard cap on rows written to the dataset.500
maxPagesMaximum requests. Each request returns up to 100 works.5

maxItems is clamped to 20,000 and maxPages to 200 per run. With the defaults, one run reads up to five pages of 100 works and stops at 500 rows, whichever comes first. A malformed filter is rejected by Crossref on the first request, which fails the run instead of returning an empty dataset.

Output dataset

FieldMeaning
titleThe work's title, taken from the first entry of Crossref's title list.
journalThe journal or container the work appeared in, from the first entry of container-title. Left out for books, datasets and other non-journal types.
publisherThe organisation Crossref records as the publisher.
authorsThe first 10 authors, each written as given name then family name, joined with ; . An eleventh author is not included.
publishedYearThe year from the work's published date, as a number. Left out when Crossref has no published date for the work.
citedByCrossref's is-referenced-by-count — how many deposited works cite this one. A citation count, not a quality measure, and only as complete as the deposits behind it.
urlThe URL Crossref records for the work, normally its DOI resolver link.

Every row also carries id (the DOI prefixed with doi:, e.g. doi:10.1038/s41586-024-00001-2), source (crossref-works), type (Crossref's work type, e.g. journal-article), firstSeenAt and scrapedAt. Works without a DOI are skipped, because the DOI is the row id. Rows are de-duplicated on the DOI and sorted by id. A SUMMARY record in the run's key-value store reports rows, requests, duplicates, page problems and whether the dataset was truncated by a cap.

Example input

{
"query": "graphene perovskite solar cell stability",
"filter": "from-pub-date:2024-01-01,type:journal-article",
"maxItems": 500,
"maxPages": 5
}

Why this is different

  • Crossref Scholarly Works uses Crossref's cursor pagination, so a long result set is read page by page instead of being cut off at a fixed offset.
  • Crossref Scholarly Works needs no key, account or registration before its first run.
  • Crossref Scholarly Works passes the filter value to Crossref unchanged, so the full filter vocabulary works without being re-implemented.
  • Crossref Scholarly Works flattens the author list into one ; -separated string, which survives a CSV export without nested JSON.
  • Crossref Scholarly Works records the DOI as the row id, so the same work cannot appear twice in one dataset.
  • Crossref Scholarly Works reports when a cap truncated the dataset instead of presenting a partial result as complete.

Limitations

  • This is metadata, not full text. Abstracts appear only when the publisher deposited one, and the body of a paper is never returned.
  • Author lists are capped at 10 names per row. Large collaborations are truncated, and the cap is not flagged per row.
  • Only the first title and the first container title are written. A work with a subtitle list or several containers loses the rest.
  • citedBy is Crossref's count. It differs from counts in other indexes, and it is not comparable across them.
  • Works without a DOI are dropped, so very old or informally published material can be missing even when it matches the query.
  • publishedYear comes from the published date only. A work with an online-first date and a later issue date reports the value Crossref holds there, which may not be the year you expect.
  • Crossref applies its own rate limits. The Actor retries a throttled or failed request up to three times with backoff; if the first request still fails, the run fails rather than returning an empty dataset.

FAQ

Do I need a Crossref API key?

No. The works endpoint used here is public and the Actor sends no credentials.

Can I narrow the search by journal or by publisher?

Yes, through filter — for example issn:0028-0836 for a journal or prefix:10.1038 for a publisher's DOI prefix. The filter is passed to Crossref unchanged, so its documented filter names apply.

Why did I get fewer rows than Max rows?

Either the result set was smaller than the cap, or maxPages stopped the run first: five requests at 100 works each is 500 rows. The SUMMARY record names the stop reason, and raising maxPages continues the search.

Is the citation count a measure of quality?

No. It is the number of works Crossref has that cite this one. It says nothing about the paper's methods or importance, and it lags behind publication.

Why is a known paper missing?

The work may have no DOI, may be outside the filter you set, or may sit beyond the page and row caps. Check the SUMMARY record before assuming the paper is not indexed.

Pricing and limits

The price is shown on this Actor's Apify pricing panel, and charges follow the pricing model displayed there. The caps in this README describe how much data the Actor requests; they are not a price. Each request returns at most 100 works, requests are paced 200 ms apart, and a run stops at maxItems rows or maxPages requests. The SUMMARY record states whether either cap truncated the result.