Datasets
The global journal directory, publishers, rankings, curated journal records and the subject classification, under open licences. Every file is regenerated from the current data.
Global index
158,455 journals registered with Crossref or OpenAlex, the 34,575 publishers behind them, and the ranking edition.
| Path | Contents | Records | Size | Formats |
|---|---|---|---|---|
| journals/index.json | Global journal directory: totals, and the file list for each subject category | 33 | JSON | |
| journals/{category}.json | Every indexed journal, one file per PSC category; categories over 4,000 journals are split by first letter | 158,455 | 33.8 MB | JSON |
| meta/publishers.json | Every publisher with indexed journals: journal, Core, open-access and DOAJ counts, and works | 34,575 | 5.0 MB | JSON |
| publishers/{00-ff}.json | Publisher details with subjects, countries and every journal, in 256 hashed shards | 34,575 | JSON | |
| rankings/citation-2026.json | Citation Ranking 2026 (PNCI-1.0): the edition's versions, snapshot date and thresholds, and the file list per subject category; the CSV has the 50,983 ranked journals | 158,242 | JSON, CSV | |
| rankings/citation-2026-all.csv | Every journal of the Citation Ranking edition, all ranking statuses, as one CSV | 158,242 | JSON | |
| rankings/citation-2026-{category}.json | PNCI, citation rank, percentile, Citation Quartile, POSI Zone and ranking status per journal, one file per PSC category, with PCS as a descriptive field and PCI, also descriptive, for Core Collection journals only | 158,242 | JSON |
Curated records
Records POSI holds with a permanent POSI-J id. Index files use short keys, documented in the record schema.
| Path | Contents | Records | Size | Formats |
|---|---|---|---|---|
| index/core.json | Core Collection records, compact index | 32 | 7 KB | JSON, CSV |
| index/benchmark.json | Global Benchmark curated seed, compact index | 993 | 257 KB | JSON, CSV |
| index/discovered.json | Discovered and other curated (not certified) records, compact index | 25,766 | 6.0 MB | JSON, CSV |
| journal/{code}.json | Full record with status and indicators, one file per Core and Benchmark journal | 1,025 | JSON | |
| records/discovered-{00-63}.json | Full Discovered records, 64 shards by a hash of the record key | 25,766 | 23.5 MB | JSON |
| meta/psc.json | PSC subject taxonomy v1.0.0 | 48 | 6 KB | JSON |
| meta/schema.json | Field dictionary and index key map (see Record schema) | JSON | ||
| meta/stats.json | Record counts by collection and verification state | JSON |
Canonical snapshot
Each snapshot is immutable and records the data and engine versions it was computed from, with SHA-256 checksums.
Published audits
Every change to the curated corpus is published with its method, counts and per-record files in the posi-data repository.
| Audit | Date | Summary |
|---|---|---|
| POSI Journal Evaluation Architecture 1.0 | 2026-09-28 | AJR Ratings (A+ to D) added from each published AJR score; legacy E-Q/M-Q quartiles, PCS-Q quartiles and PCI Citation Q kept as archive and retired as rankings. Citation Rankings move to PNCI-1.0. |
| Global index and PCS edition | 2026-09-27 | Every journal registered with Crossref or described by OpenAlex, 158,242 after merging on ISSN; PCS computed from Crossref for each (journals looked up under each of their ISSNs). Its PCS-Q ranks are retired (POSI-EVAL-1.0). |
| PCS ETL, full scope | 2026-08-14 | POSI Citation Score from Crossref data for 4,320 journals (the 31 Core Collection journals and the full 4,289-journal Global Benchmark): 6.77 million works fetched, PCS computed for 4,089. |
| AJR-E-1.1 rerate, Core Collection | 2026-08-14 | First run of the Early-Stage rating against all 31 Core Collection journals, combining site-crawl and Crossref article-sample evidence. |
| Publisher-expansion canonical records | 2026-08-13 | Wrote the 2,177 journal records that the Elsevier and Frontiers expansion had minted ids for, so every registry id resolves to a record. |
| Citation preview correction | 2026-08-13 | Withdrew provisional citation quartiles on 3,245 Global Benchmark records and replaced them with a diagnostic-only preview: Benchmark membership alone does not make a journal ranking-eligible. |
| Elsevier and Frontiers expansion | 2026-08-12 | Global Benchmark grown from 1,000 to 4,289 records from the publishers' own title lists (3,113 Elsevier, 183 Frontiers), with every identity conflict resolved in a second-round re-run. |
| Core and Benchmark identity remap | 2026-08 | All 1,000 Global Benchmark and 31 Core Collection journals resolved against the 24,205-record registry: 0 new ids, 0 conflicts, 0 left for manual review. |
| Initial journal migration | 2026 | 23,822 legacy source records audited in two independent dry runs with byte-identical output; 171 possible-duplicate groups resolved (166 merged, 5 kept distinct); 23,331 POSI-J ids minted, 0 collisions. |
Licences
| POSI-produced data (PSC, scores, rankings, curated metadata) | CC BY 4.0 |
| Source code of this site and posi-engine | MIT |
| Upstream metadata (Crossref, OpenAlex, DOAJ, ROR, ORCID) | Source license |
POSI is open source. Third-party open data are credited and keep their original licences.
Cite the dataset
Panorama Open Scholarly Index (2026). POSI journal records. Panorama Scholarly Group. https://posi.panorama-sg.com/datasets/. Accessed [date]. Licence: CC BY 4.0.