Due to different manufacturers and internal, non-standardized data structures, media workflows are often not interoperable. As a result, a lot of information is lost in the course of the production and distribution process and must be aggregated across systems with high personnel expenditure. The interconnection of the value chain and the structuring of the information should therefore be improved by not only generating unique metadata records but also capturing their significance.
Semantically networked metadata enable novel assistance systems for editors and journalists for intuitive semantic search and navigation. It also makes it possible to link the results from audio-visual analysis tools (speech-to-text etc.) with semantic text structuring solutions for the (partially) automated enrichment of metadata down to scene level (e.g. "In which scene does Angela Merkel shake hands with Vladimir Putin?").
End users also benefit from the semantic networking of metadata, e.g. through more intuitive search and navigation options in media libraries or apps. This also enables more efficient semantic recommendations in recommender systems based on user behavior.
In cooperation with broadcasters, IRT is testing fields of application for semantics with the aim of offering innovative media services and enabling savings in the various areas of the media value chain. In addition to technical aspects, economic, qualitative, structural, organizational, competitive and standard relevant aspects are also considered.
More about this