I am looking for a public corpus of dialogue recording in short/medium appointment between doctors and patients, aims to study speech variation, the prosody of speech, measuring speech' rhythm, innovation and so on. I need relatively large audio recording sample, so I can train those on the specific neural network to precisely predict and classify the speech by different specific criteria.

I have searched from possible NLP labs around but couldn't find the corpus that meets my expected research. Is there any experienced linguistics experts could point me how to get possible corpus for NLP research? Can anyone guide me how to acquire the corpus that I stated above? Thanks in advance

secondary preferable corpus

If getting above corpus is not feasible at this moment, I'd like to try other spoken recording corpus with the following specification:

  1. corpus of monologue recording in reading newspaper/short stories, poem/reading basic sentence (English, Italian, or other European language can be taken into consideration)
  2. corpus of the audio recording in conversation/dialogue between standard English/Italian (or other European languages) and its specific dialect counterpart.
  3. or other language corpuses (spoken) that closely aligned the statement that I mentioned above.
  • 4
    I am sure you have heard about patient confidentiality.
    – fdb
    Commented Jan 17, 2018 at 22:30
  • I'm voting to close this question as off-topic because this question is not about Linguistics or NLP. Questions about finding datasets may better fit at Open Data. Namely, they have tags medical and audio which you may find useful. VTC/offtopic. Commented Jan 17, 2018 at 23:47
  • Richard Frankel, a sociolinguist, has done a very large amount of research on how doctors an patients interact.
    – jlawler
    Commented Jan 18, 2018 at 0:16
  • Because of confidentiality, voice recordings could not be made public. Anonymized transcriptions would be allowed.
    – user6726
    Commented Jan 18, 2018 at 2:01
  • @user6726 Would it possible to acquire corpus that I stated above if I provide my research center' formal request and credential? Could you guide me the possible route I could give it try? Thanks
    – Andy.Jian
    Commented Jan 18, 2018 at 10:47

1 Answer 1


Contrary to the expectations of some commentators, doctor-patient corpora are available (under some conditions, needing to sign some licence and confidentially agreement) for research. The standard entry point for a search for such corpora is the CLARIN Virtual Language Observatory and entering doctor patient in the search slit gives currently twelve results.

Here is a bookmarked query

  • Thank you so much for your help, this is definitely breathtaking for me to further explore.
    – Andy.Jian
    Commented Jan 18, 2018 at 10:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.