About the Dataset

859 hours

This audio dataset contains 859 hours of Spanish (Mexico) Speech Data in various domains, recorded by speakers from Spain, Mexico, and the US.

There are 438 hours of Spanish (Spain) human-to-human audio with the following domain distribution per dataset:

• 88.55 hours of Banking
• 60.1 hours of Insurance
• 71.45 hours of Retail
• 218.83 hours of Telecommunication

There are 318 hours of Spanish (Mexico) human-to-human audio with the following domain distribution per dataset:

• 87.23 hours of Banking
• 92.15 hours of Insurance
• 74.93 hours of Retail
• 63.97 hours of Telecommunication

This audio dataset contains 103 hours of Spanish (US) Speech Data in generic domain, recorded by speakers from the US.

Defined.ai creates scenarios for our crowd members to follow, which they study beforehand. They then record a conversation, one speaker playing the agent, the other speaker “playing out” the scenario with spontaneous content. The recording is done via telephony and is saved in 8kHz 16 bit per channel. That content is then transcribed.

The dataset is covered by Defined.ai's standard license agreement. The license agreement is perpetual and allows for the commercialization of all models built on the data.

Other characteristics:

• Audio format: WAV
• Recording environment: noisy, silent
• Bits per sample: 16
• Communication band: broadband
• Sample rate: 8kHz

Metadata Distribution

Spanish (Spain)

Spanish (Mexico)

Samples

5-minute audio samples

• Spanish (Spain) Banking. Transcription for the sample is also available
• Spanish (Mexico) Banking. Transcription for the sample is also available
• Spanish US Generic Domain. Transcription for the sample is also available