Hugging Face's logo

Datasets:
KBLab
/
rixvox

Tasks:
Automatic Speech Recognition
Modalities:
Audio
Languages:
Swedish
Size:
100K<n<1M
Tags:
audio
speech-recognition
License:
Dataset card Files Community
3
rixvox / README.md
Lauler's picture
Lauler
Metadata fields info
8ad28ec
preview code
|
raw
history blame
2.41 kB
metadata
language: sv
license: cc-by-4.0
tags:
  - audio
  - speech-recognition
task_categories:
  - automatic-speech-recognition

RixVox

A Swedish speech corpus with 5500 hours of speeches from debates in the Riksdag (the Swedish Parliament). The speeches have been aligned with transcripts from written protocols.

Metadata

Guide for metadata fields:

  • dokid: Document id for the debate. This is the same for all speeches in a debate.
  • anforande_nummer: Speech number within the debate, or within the debate sessions on a particular day. Should create a unique primary key in combination together with dokid (sometimes there are duplicates, but we removed them from this dataset).
  • start: The start time of the observation within a speech (in seconds).
  • end: The end time of the observation within a speech (in seconds).
  • duration: The duration of the observation (end subtracted with start).
  • intressent_id: Unique id for the speaker within the Riksdag's database (see person.csv.zip from the Riksdag).
  • speaker: The speaker's name retrieved via the intressent_id.
  • party: The speaker's party retrieved via the intressent_id.
  • gender: The speaker's gender retrieved via the intressent_id.
  • birth_year: The speaker's bith year retrieved via the intressent_id.
  • electoral_district: The electoral district which the speaker represents if they are/were a member of parliament (retrieved via the intressent_id).
  • speaker_audio_meta: The speaker's name and title as listed in the Riksdag's oroginal text format metadata (sometimes wrong and mismatched against intressent_id).
  • speaker_from_id: Whether the speaker metadata was retrieved via the intressent_id or via the Riksdag's original metadata, for those speeches with a missing intressent_id.
  • is_prediction: We reverted back to the Riksdag's metadata for a small subset of modern speeches (500 out of 117k). A True value of is_prediction indicates that the speech was predicted by KBLab. A False value indicates that we used the Riksdag's original metadata.
  • debatedate: The date of the debate.
  • bleu_score: The BLEU score of the automatic speech recognition (ASR) transcript against the Riksdag's written protocol. Calculated on the entirity of the speech that an observation (30s snippet) is extracted from.