SNOMED CT Entity Linking Benchmark
SNOMED International / DrivenData
Overview
Model evaluation linking clinical-note terms to SNOMED CT concepts.
Included as a machine learning/algorithm task; generative AI use is not assumed to be mandatory.
Notice dates
Calendar dates from the saved notice. They do not establish a deadline time, time zone or current acceptance status.
Eligibility
Individuals or teams; comply with applicable law, employer policies, and data-access conditions.
Cautions and verification notes
This is a separate, continuously available benchmark from the earlier SNOMED prize competition; it is not a duplicate.
Closed-source models are allowed. Access to actual clinical data requires a separate procedure.
The official overview and body text were checked, including the overview and embedded benchmark rules. Eligibility for separate PhysioNet data access remains unconfirmed.
These requirements concern the AI/ML task and do not establish blanket permission to use generative AI assistance.
Sources
Check the original notice for current rules. English includes editorial renderings of stored notices and organizer names, not verified official English names. Original records are preserved. Excerpts may be incomplete; consult the linked sources for full conditions.