Currently metadata is created for every audio, including task instructions and prompts. This metadata is then deidentified the same ways audios are meaning that they are removed currently for free speech and not added to bundles.
Options to consider:
- Leave as is since metadata is related to audios themselves and shouldn't be included if an audio isn't going to be included
- Treat as a static feature which is released no matter what since none of the information is identifiable
- Create a static document of audio instructions/prompts to release for the bundle since that information could be stored more compactly as a static set (would need to decide between above for remaining metadata)
Currently metadata is created for every audio, including task instructions and prompts. This metadata is then deidentified the same ways audios are meaning that they are removed currently for free speech and not added to bundles.
Options to consider: