The AI4LAM Speech-to-Text Working Group is pleased to announce version 1.0 of Transcript Provenance Metadata Elements (TPME), published on July 10, 2026.
TPME is a set of metadata elements designed to record the provenance of common informational artifacts (like transcripts) that are derived from primary media items, often by way of AI systems.
The schema was developed over 18 months through an open drafting and revision process involving practitioners across libraries and archives.
TPME is published as both PDF (for use by humans) and Markdown (for use by humans or computers), available here: https://doi.org/10.5281/zenodo.21287085
To provide feedback, please comment on the current working version: https://docs.google.com/document/d/12cKw5EX9odQr0fZBG3Emkb1in5k1nQvu33dmSwbr6FQ
Kind regards,
Owen King, and the rest of the AI4LAM Speech-to-Text Working Group
Owen King
Metadata Operations Manager
Received on Sun Jul 26 2026 - 22:30:58 EDT