Spontaneous Dialogue

Dutch Speech Data - Spontaneous Dialogue

Insurance

Version Number

Published date

Sep 1, 2021

Audio demo

We can collect a customized dataset according to your precise needs.

call centre, conversational AI

Acoustic Modelling, ASR Testing, Benchmarking

A zip file containing metadata files in tsv format and a folder with all the audio files

This dataset contains 420 hours of Dutch Spontaneous Dialogue data, recorded from speakers in Netherlands.

Defined.ai

Dataset details

Domain	Insurance
Total recordings	8591
File size	45.68GB
Hours	420
Word error rate (%) Measurement indicating errors in alignment of text representation (actual vs. perfect) of audio, taking into account words omitted, inserted or wrongly replaced.	3.4%
Average amount of recordings per speaker	116.09

Locale The language(s) and country(s) applicable to the speakers in the dataset.	nl-nl
Language	Dutch
Country	Netherlands
Female \| Male \| Unspecified View on chart	54% \| 45% \| 1%
Age View on chart	18-62
Accent(s) View on chart	ABN, Friesland, Gelderland, Groningen (Gronings), Limburg (Limburgs), Noord Brabant (Brabants), Noord Holland, Overijsel, Zuid Holland