Preprint
PriMock57: A Dataset Of Primary Care Mock Consultations
arXiv, 2022
Read the originalRead on October 16, 2026
- What it is
- The abstract of the PriMock57 paper, which describes a public dataset of mocked primary care consultations with audio recordings, manual utterance-level transcriptions and consultation notes. The authors built it because clinical datasets are hard to share.
- How much weight it carries
- An arXiv preprint whose abstract we read directly. The consultations are mock sessions, not real patient visits, so the dataset tests recognition on realistic conversations but does not show how real clinics sound.
Facts taken from this source
- Access to clinical datasets is heavily restricted because of patient privacy, which is why the PriMock57 authors built a public mock dataset.
“access to clinical datasets is heavily restricted due to patient privacy” (Abstract, first two sentences)
- PriMock57 is a public dataset of mocked primary care consultations, with audio recordings, manual utterance-level transcriptions and consultation notes.
“mocked primary care consultations, including audio recordings, their manual utterance-level transcriptions, and the associated consultation notes” (Abstract)