The steady growth, year-on-year, of the EBU Metadata Developer Network Workshop points to the increasingly central role of data and AI for broadcasters. This year's MDN Workshop, running across three days for the first time, has brought more than 80 attendees to Geneva. (The event continues until tomorrow; presentations (and later videos) are available to anyone with an EBU website user account.)
On the opening day, several speakers explained why data is key to improving the quality of service proposed to the audience. Concepts like “knowledge management”, “data governance”, “information architecture” and “semantic organization” are now well understood. Good data sustains public service media. As EBU Members have to do more with less, it is important to optimize working structures and attract key talent, like system architects and data scientists, to make the best out of what new technologies propose around metadata.
Cheaper data via AI
The use of AI in production continues to grow, particularly for automatic metadata extraction. Its use for tasks such as scene detection, face recognition and genre (e.g. sport) detection generates cheaper metadata. AI means more data, which means better data management is also required.
Having increased and richer data makes it easier for broadcasters to find and re-purpose/re-publish content across platforms. However, AI integration into operational workflows is an essential implementation step using service-based architectures.
Several EBU projects like EBUCore, the Class Conceptual Data Model (CCDM) and MCMA (Media Cloud and Microservice Architectures), have been mentioned in presentations.
Metadata experts and implementers are encouraged to join the Metadata Developer Network, a community to collect and share best practices.
Clear evidence of how interest has grown around metadata in media!