Work repo for Stortinget Speech Corpus

Short introduction

This repo contains the code used for creating the Stortinget Speech Corpus, a large speech corpus with speech from Stortinget and transcriptions extracted from the proceedings at Stortinget. See our paper for more information about this dataset.

Content

• data/ contains ASR transcriptions and proceedings
• matching/ contains a modified version of the matching code from CLARINSI.

Matching

The ASR transcriptions need to be inverse-normalized. Clone and install the normalization code:

git clone https://github.com/Sprakbanken/sprakbanken_normalizer.git

python -m pip install .

Example code

from sprakbanken_normalizer.inverse_text_normalizer import inv_normalize

print(inv_normalize("dette tallet er tre hundre tusen fire hundre og tjueto"))

Saving results to an SQLite database

An SQL database needs to be built first to save the references to the proceedings and the transcriptions.

python3 run.py results.db

Make a csv file with extracted transcriptions

python3 make_match_csv.py /path/to/outfile.csv