MedWER — provenance and attribution
====================================

Code
----
Copyright (c) 2026 NORDIS TECH INC. Licensed under the MIT License (see LICENSE).

The text normalizer implements the pipeline of Whisper's EnglishTextNormalizer
(OpenAI, MIT License), consumed via the whisper-normalizer package
(https://github.com/kurianbenoy/whisper_normalizer, MIT License).
medwer/data/english.json is the GB->US spelling map exported from that package.

Term list (medwer/data/medical_terms.txt)
------------------------------------------
Drug terms:
  Contains information licensed under the Open Government Licence – Canada.
  Sources: Health Canada Drug Product Database;
  Canadian Clinical Drug Data Set (Canada Health Infoway / Health Canada),
  retrieved from the public Health Canada distribution.
  https://open.canada.ca/en/open-government-licence-canada

Diagnosis / symptom / mechanism terms:
  ICD-10-CM (National Center for Health Statistics / Centers for Medicare &
  Medicaid Services). United States public domain.

Organism and laboratory terms:
  Original curation by NORDIS TECH INC., licensed MIT.

The list contains bare term strings only. It encodes no drug-name confusability
relationships and no data selected or compiled by third parties beyond the
sources above.
