OpenAlex Romania: scholarly works and institutions

The Romanian slice of OpenAlex, the free open catalogue of the global research system (built by OurResearch, successor to Microsoft Academic Graph). Institutions located in Romania, and every scholarly work OpenAlex indexes with at least one Romanian-affiliated authorship — article, book chapter, preprint, dataset. Authorships are kept relational rather than nested: work_authorships is one row per (work, author, Romanian institution), so a reader can ask "who are Romania's most-cited researchers" or "which field is Romania publishing most in" directly in SQL. REFERENCED LIVE, NOT COPIED: all three tables are published as views over OpenAlex's own Parquet snapshot on the AWS Open Data bucket s3://openalex, so no row is stored in this directory's lake and every query reads OpenAlex directly — a citation count OpenAlex has since revised, or a correction to a title or a retraction flag, shows up in the next query with no re-ingest here. That also means there is no history beyond the one query each view was published with, no row counts or column statistics of our own, and a query needs network access to s3://openalex (public, no credentials). The bucket has no partition on country, so filtering works/work_authorships to Romania costs a genuine column scan of the works archive — this connector publishes the complete archive's view by default (since=all), which costs nothing to publish under federation; a query against the resulting view pays whatever scan the query itself needs.

referenced liveRows

3Tables

—Size

10 hours agoLast ingest

CC0-1.0Creative Commons Zero v1.0 Universal (public domain dedication)

Attribution, terms and provenanceWhat you must carry if you reuse this

Attribution: Data source: OpenAlex (https://openalex.org).
Notes: OpenAlex's own licence page states plainly: "OpenAlex data is made available under the CC0 license. That means it's in the public domain, and free to use in any way you like," adding that attribution "is appreciated where it's convenient, but it's not at all necessary." Verified directly against github.com/ourresearch/openalex-docs/blob/main/license.md on 2026-09-02. The one carve-out OpenAlex documents is its legacy "MAG Format" export, released under ODC-BY because it inherits Microsoft Academic Graph's own licence — this connector does not use that export, only the current Parquet snapshot, so ODC-BY does not apply here. IMPORTANT CAVEATS FOR A REUSER, none of which change the licence but all of which change how far to trust a value. (1) OpenAlex itself aggregates many upstream sources (Crossref, PubMed, ORCID, ROR, institutional repositories, publisher metadata) and inherits their coverage gaps and occasional errors — a missing DOI or an unresolved affiliation is usually an upstream gap, not evidence the work does not exist. (2) Affiliation-to-institution matching is automated: a work with institutions_distinct_count = 0 has an author who wrote something OpenAlex could not resolve to a canonical institution, and is correctly excluded here rather than guessed at. (3) cited_by_count, fwci and works_count grow over time as OpenAlex indexes new citing works — and because works/institutions are published as VIEWS over OpenAlex's live bucket (not copied), a later query naturally sees the larger, current number with no re-ingest of ours involved: two queries run minutes apart can legitimately disagree, and the later one is the more current value, not a data-quality problem. (4) What is published here is a filtered, reshaped VIEW over OpenAlex's own data — the Romanian slice, presented as relational tables — not a stored copy and not a verbatim republication of the OpenAlex snapshot; the SELECT that shapes it is published in full as each table's view definition.
Version: snapshot 2026-10-22 · 01M5GSRHR2T9X2A4ZJ159GK0HA
Machine-readable:: llms.txt manifest.json

Tables

Open a table for its columns

Loading the schema…

Queries

Copy a snippet, or run it in the playground

Query it yourselfConnection snippets for DuckDB — Parquet or the live lake

Generated from this exact manifest version, so the URLs are the ones actually published.

Attach the catalog, then query it

CREATE OR REPLACE SECRET "openalex_pds" (TYPE s3, SCOPE 's3://openalex', PROVIDER 'config', REGION 'us-east-1');

ATTACH IF NOT EXISTS ':memory:' AS "openalex_ro";

CREATE SCHEMA IF NOT EXISTS "openalex_ro"."main";

CREATE OR REPLACE VIEW "openalex_ro"."main"."institutions" AS SELECT
  id                                    AS institution_id,
  ror,
  display_name,
  lower(strip_accents(display_name))    AS name_lc,
  "type"                                AS institution_type,
  country_code,
  homepage_url,
  works_count,
  cited_by_count,
  geo.city                              AS city,
  geo.region                            AS region,
  geo.latitude                          AS latitude,
  geo.longitude                         AS longitude,
  DATE '2026-10-22'                     AS first_seen_date,
  DATE '2026-10-22'                     AS last_seen_date,
  TIMESTAMPTZ '2026-10-22T03:16:34.178Z'                            AS ingested_at
FROM read_parquet('s3://openalex/data/parquet/institutions/**/*.parquet', hive_partitioning = true, union_by_name = true)
WHERE country_code = 'RO';

SELECT *
FROM openalex_ro.main.institutions
LIMIT 10;

Execute