Catalog

German Spontaneous Dialogue data - 509 hours

Overview

Banking, Insurance, Retail, Telecommunication
Domain The business or industry verticals that characterize this dataset
7.04%
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.
de-de
Locale The language(s) and country(s) applicable to the speakers in the dataset.
18-66
Age Age range of the speakers in the dataset.
43% | 52% | 5%
Female | Male | Unspecified Percentage of female and male speakers in the dataset.
$216,325.00
Version Number
01
Hours
509
Purchase Options
Sender
Invitee
Apr 8, 2021
Published date Date the dataset was published.
DefinedData
Seller Name Entity or individual making the dataset available for sale.
DefinedData
Data source Source entity providing the dataset
License type Agreement defining the data being licensed, including the usage, manner, and frequency in which data will be provided or updated.

Use case(s)

call centre, conversational AI

Model Applications

Acoustic Modelling, ASR Testing, Benchmarking

Packaging description

A zip file containing metadata files in tsv format and a folder with all the audio files
This dataset contains 509 hours of German Spontaneous Dialogue data, recorded from speakers in Germany.

Dataset details

About

Domain The business or industry verticals that characterize this dataset Banking, Insurance, Retail, Telecommunication
Total recordings Total audio recordings in the dataset. 7846
File size Dataset file size. 54.86GB
Hours Total number of hours in the dataset. 509
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. 7.04%
Average amount of recordings per speaker Average amount of audio recordings per speaker in the dataset. 29.89

Demographic

Number of speakers Total number of unique speakers in the dataset. 525
Locale The language(s) and country(s) applicable to the speakers in the dataset. de-de
Language Language(s) spoken in the audio files included in the dataset. German
Country The country where the speakers reside and are represented in the audio. Germany
Female | Male | Unspecified Percentage of female and male speakers in the dataset. 42.67% | 52.38% | 4.95%
Accent(s) Accents spoken by speakers in the dataset. Alemannisch (Stuttgart, Ulm), Bayerisch (München, Nürnberg), Ostmitteldeutsch (Dresden, Leipzig, Berlin), Westmitteldeutsch (Köln, Bonn, Frankfurt), Westniederdeutsch (Hamburg, Bremen, Hanover)

Age Distribution for Agents

Age Distribution for Agents
Age Distribution for Agents

Age Distribution for Customers

Age Distribution for Customers
Age Distribution for Customers

Gender Distribution for Agents

Gender Distribution for Agents
Gender Distribution for Agents

Gender Distribution for Customers

Gender Distribution for Customers
Gender Distribution for Customers

Audio details Additional audio information.

  • Words
    4111677
  • Recording environment Acoustic environment in which the recordings were made.
    noisy, silent
  • Audio format File format of audio in the dataset.
    WAV
  • Bits per sample Number of bits used to represent the audio data for each sample of playback.
    8
  • Device type Type of device the audio was recorded on.
    mobile
  • Communication band Channel bandwidth used during the recording of the audio.
    narrowband
  • Sample rate Number of samples per second for audio data.
    8kHz

Sample

Sample demo Audio clips from the dataset that you can listen to.
Name Duration Age Gender Native Language
1 Audio sample 1 00:03:28.680 N/A N/A de-de