There is a growing realization that sound data architectures are essential to serving modern media audiences. We asked three of those active in the area to describe how (meta)data is handled where they work.
Jürgen Grupp, Information Architect, SWR
My role involves advising projects on semantical and technical aspects. The first step is always about finding a common language and a common view of the business. The quality of communication is decisive for the quality of the solutions we find. By having the right terms and definitions in place, we avoid expensive detours and inappropriate solutions. Sound information architecture is the stable and flexible infrastructure of any digital business process.
Most people in the organization value data as indispensable for their daily work. We have identified the need for building a strategy and coordinating the various initiatives. A data board has been established, bringing together business and technology views from across the organization. The board will prioritize, direct and initiate projects to create business value.
As chair of the Editorial Committee for EBUCorePlus, the semantic information model developed and maintained by the EBU, I see great benefit in bringing Members together and defining a common semantic model. This will speed up system integration, reduce vendor lock-in for your own organization and, in the long run, foster data and metadata exchange between Member organizations.
Samuel Profumo, Chief Data Officer, RTBF
I oversee the development and implementation of public service algorithms, ensuring they adhere to our core values of ethics, transparency and alignment with our public mission. My role also involves developing advanced audience analytics, ensuring the precise management and organization of metadata, and leading the process of automatic content enrichment to enhance the value and relevance of our media offerings.
RTBF’s strategic vision RTBF27 sees data as a fundamental pillar, aiming to “bridge the gap between citizens living in a personalized world”. Our recent shift towards comprehensive metadata management, epitomized by the development of a unified content management portal, marks a significant step in this direction. The industrialization of automatic metadata extraction processes through AI is also a key driver in our transformation. Now engaging with most of our audience through registered profiles on our digital platforms, we have a robust foundation for personalizing and enriching their experience.
The wealth and quality of our trusted content provide fertile ground for joint innovation at EBU level. Pooling and leveraging our data can spearhead the development of new AI tools, facilitating the work of our editorial teams and enhancing the audience experience. The idea of creating shared data platforms and establishing standardized protocols for data management and privacy is particularly promising.
Jeremy Tarling, Head of Content Metadata, BBC
I am responsible for the development, coverage and quality of content metadata for BBC’s digital product portfolio, including iPlayer, Sounds and BBC’s various websites and apps including News, Sport, Bitesize (education) and children’s content.
As a 100-year-old broadcaster, BBC has many established ways of working based around linear television and radio production. A key part of our long journey (the first BBC website went live in December 1997) towards a truly “digital-first” BBC is our focus on data-driven decision making, from programme commissioning through to the audience experience of our digital products. Rich content metadata descriptions of programmes and other digital content, combined with audience consumption data, helps us to build a more personal BBC, using machine learning and AI to drive algorithmic experiences of a BBC that feels more like it is ‘for me’.
The EBU has been instrumental in the development and governance of technical standards that are a constant reference in our digital transformation. Participation in the EBU’s data workstreams has been very helpful for knowledge sharing and common problem solving – for example in the Metadata and AI group that covers metadata standards like EBUCorePlus together with challenges like AI benchmarking and automatic metadata generation.
This article first appeared in issue 59 of EBU tech-i magazine.