metadata

language: id
license: gpl-3.0
task_categories:
  - automatic-speech-recognition
tags:
  - medical
dataset_info:
  splits:
    - name: train
      num_bytes: 2705299088
      num_examples: 360
    - name: test
      num_bytes: 479374176
      num_examples: 360
  download_size: 3184673264

Building MEDISCO: Indonesian Speech Corpus for Medical Domain

The dataset was published in the following paper:

Building MEDISCO: Indonesian Speech Corpus for Medical Domain (PDF | IEEEXplore)
Muhammad Reza Qorib and Mirna Adriani
2018 International Conference on Asian Language Processing (IALP)

Please look for the raw files (inside the "Files and versions" tab) as the dataset viewer parsed by Huggingface does not show the text transcript. Please direct any questions to mrqorib[at]u.nus.edu as the email addresses in the paper PDF are no longer active.