MétaCan
ENFR
Menu
Cohort builder

4,299,418 works, Canadian by any of four routes.

Every filter state is a URL; the URL is the query; the query is citable via /q/⟨hash⟩. The page, the API and the export parse the same parameters.

Export CSVExport JSONAPI query

The current cohort, streamed from the database: every work column, the machine labels, the provisional scores, and the per-row validation status. Exports are capped at 100,000 rows. Mints a permanent /q/ link for this exact query. The same filters always produce the same link, whoever asks.

Search term
Author
Year range
→
Sort
Language
Type
Field
Venue
Topic
Topic Modeling
Retraction
Abstract
Evidence source
Study design
Label agreement
Label status

Direct Codex and Gemma labels are unvalidated and sparse. Distilled predictions cover the full frame and are also unvalidated. Choose the evidence source explicitly; absence of a direct label is never a negative label.

affaffiliation
fundfunder
venuejournal
aboutaboutness

The four routes compose: require the funder route and exclude affiliation to get the funder-only stratum no affiliation-based frame ever sees.

2,769 results · 1 filter active ·
Results by year
20002025
Publication date
Categories
Machine labels · sparse coverage
Evidence
Language
Type
Citations
An unlabeled work is unknown, not a negative. Label coverage is reported on every query.
2,769 works in the cohort · of 4,299,418page 30 of 56

Labels cover 6 of 2,769 works in this cohort. The rest are unlabeled, which is not a negative label: the label table is sparse today and grows as labeling rounds land.

Distilled predictions cover 2,769 of 2,769 works in this cohort. Predictions are machine_predicted_unvalidated teacher distillation outputs. Candidate is the union; consensus is the intersection.

afffundunlabeled
Language Modelling via Learning to Rank
Arvid Frydenlund, Gagandeep Singh, Frank Rudzicz
2022· article· en· Proceedings of the AAAI Conference on Artificial Intelligence· Computer Science
distilled prediction:candidate · noneconsensus · none
3
citations
affno abstractunlabeled
WASM: A Dataset for Hashtag Recommendation for Arabic Tweets
Maged S. Al-Shaibani, Hamzah Luqman, Abdulaziz S. Al-Ghofaily, Abdullatif Alnajim
2024· article· en· Arabian Journal for Science and Engineering· Computer Science
distilled prediction:candidate · noneconsensus · none
3
citations
afffundunlabeled
Few shot learning approaches to essay scoring
Robert K. Helmeczi, Savas Yıldırım, Mücahit Çevik, Sojin Lee
2023· article· en· Computer Science
distilled prediction:candidate · noneconsensus · none
3
citations
affunlabeled
Automatic Answer Typing for How-Questions
Christopher Pinchak, Shane Bergsma
2007· article· en· North American Chapter of the Association for Computational Linguistics· Computer Science
distilled prediction:candidate · noneconsensus · none
3
citations
affno abstractunlabeled
Deep Learning for Document Representation
Mehran Kamkarhaghighi, Eren Gultepe, Masoud Makrehchi
2019· book-chapter· en· Smart innovation, systems and technologies· Computer Science
distilled prediction:candidate · noneconsensus · none
3
citations
venueno affunlabeled
Stanford's 2013 KBP System
Gabor Angeli, Arun Tejasvi Chaganty, Anne Lynn S. Chang, Kevin Reschke, Julie Tibshirani, Jean Y. Wu +3 more
2013· article· en· Theory and applications of categories· Computer Science
distilled prediction:candidate · noneconsensus · none
3
citations
← previousnext →

How this was built: Screen · Findings · About