Data coverage
Publication indexing and affiliation information vary by venue and year. A successful API fetch establishes which records the source returned; it does not establish that every published paper was indexed.
Loading release information…
Collection coverage by area
“Refreshed” counts files with completed source queries or checked enrichment during this run. For Machine Learning, the existing proceedings list is retained and paper-specific affiliations are recovered where supported. “Retained” files use earlier source observations. Records include non-research items subsequently excluded from ranking. Affiliation signal means an author has an institution or raw affiliation string; it does not certify correct matching or education credit.
| Area | Refreshed files | Retained files | Records | Affiliation signal |
|---|
Scopes without a complete fresh observation
These scopes have no accepted fresh source observation, including current-year editions whose exact proceedings source is unverified. Earlier cached observations are retained where available; a missing source observation is not evidence of zero publications.
| Venue | Year | Collection result |
|---|
Limits
- The current calendar year is partial. Earlier years can also have incomplete indexing or affiliation data.
- Machine Learning remains off by default. Proceedings membership and coverage are not fully verified. Paper-specific affiliations are used where recoverable; author-profile affiliations are withheld from historical credit.
- Institution identity, branch-campus classification and duplicate provider records remain sources of uncertainty. Corrections are recorded separately from source observations.
- Institutions with unresolved geography remain in World results and are omitted from country and region filters.
Download the release manifest for file hashes, validation results and the complete venue-year inventory. Project source is maintained on GitHub.