Norwegian Parliamentary Speech Corpus (NPSC)

Provider
Norwegian Language Bank (Sprakbanken), National Library of Norway
License
CC-BY-SA-3.0 (source)
Size
130 files
Format
JSON (in sampled files)
Stored at
Hugging Face (https://huggingface.co/datasets/NbAiLab/NPSC)
Access
Free and public. No account with the data provider is needed.
Category
Speech & Audio
Source
https://www.nb.no/sprakbanken/ressurskatalog/oai-nb-no-sbr-58/

About Norwegian Parliamentary Speech Corpus (NPSC)

The Norwegian Parliamentary Speech Corpus (NPSC) is a speech corpus made by the Norwegian Language Bank at the National Library of Norway in 2019-2021. It consists of recordings of speech from Stortinget, the Norwegian parliament, with orthographic transcriptions in both Norwegian Bokmal and Norwegian Nynorsk. All transcriptions were made manually by trained linguists or philologists and then proofread for consistency; entire days of parliamentary meetings are transcribed. It is a primary resource for Norwegian automatic speech recognition and for research on spoken Norwegian and its two written standards. This repository holds one JSON transcript file plus 16 kHz and 48 kHz mp3 tar archives per meeting day under data/train, data/eval and data/test (data/train lists 32 meeting days, about 6.7 GB; measured 2026-11-14; the Hub reports about 82 GB) and a loading script. Load with datasets.load_dataset('NbAiLab/NPSC') or read the JSON and audio archives directly. The official release with word-level and speaker metadata is in the Language Bank's catalogue. Sound and transcriptions are released under CC0; the Hugging Face curation is CC BY-SA 3.0.

How to load Norwegian Parliamentary Speech Corpus (NPSC) in Python

from huggingface_hub import hf_hub_download
path = hf_hub_download(repo_id="NbAiLab/NPSC", filename="NPSC.py", repo_type="dataset")
print(path)

Data structure

File types in sampled files: JSON

Files

Example file: NPSC.py

Analyze this dataset without downloading it

On Scigantic, Norwegian Parliamentary Speech Corpus (NPSC) opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.

This page is also available as Markdown and as JSON.

Related Speech & Audio datasets