TonyFANgr/localaiiaasr

Emergency Room FAQ ASR Evaluation Set

Small audio evaluation set for testing automatic speech recognition (ASR) on emergency-room FAQ questions, recorded in both Taiwanese Hokkien (nan) and Mandarin (cmn). Each question is spoken in both languages, so every transcription has a matching pair of audio clips.

The questions come from eval/FAQ_dataset in the local_aiia project, grouped into four categories that mirror the original hospital FAQ collection:

category	source file	# questions
waiting	1waiting.json	34
examination	2examination.json	22
observation	3observation.json	48
other	4other.json	29

Structure

{language}/{category}/id_{id}_{english_slug}.wav

• language: Hokkien or Mandarin
• category: 1waiting, 2examination, 3observation, 4other
• id: matches the id field in the corresponding eval/FAQ_dataset/*.json file
• english_slug: short English gist of the question, for readability only — the authoritative transcription is in metadata.csv

metadata.csv follows the Hugging Face AudioFolder convention (file_name column) so the dataset loads directly with:

from datasets import load_dataset
ds = load_dataset("TonyFANgr/localaiiaasr")

Columns: file_name, transcription (original Traditional Chinese text), language, category, id.

Notes

• Total: 266 clips (133 questions × 2 languages).
• Intended for small-scale ASR sanity checks / regression testing, not large-scale training.