Within the Flemish media sector, metadata and its exchange between companies is considered a hot potato. A common thesaurus would have a lot of added value for the sector by highly harmonizing the understanding of proper names and concepts. But the tools to extract these names and concepts from texts as automatically as possible are, of course, important as well.

For the Flemish media companies, the general aim of this sector initiative is to investigate methods and systems that can classify all types of media items (texts, sounds, images) to better disclose them and launch new, personalised media products in a multi-company context.

This study will take the opportunity to tackle the challenge of a common, distributed thesaurus with the archives, heritage and media sector to reach a broadly supported solution.

At the start of the Flemish Institute for Digital Archiving and the disclosing of Audiovisual Heritage (VIAA), the importance of metadata and accessibility was always highlighted. Of course, VIAA will take initiatives to describe all VIAA-stored content as well as possible, by using existing metadata or creating new annotations (automatically or manually if needed) in collaboration with the content owners. This would however be a lot easier to realise if the incoming metadata was already harmonised to a certain degree. The Unified Thesaurus would tackle precisely this challenge and thus provide VIAA with greater efficiency in annotating and added value in terms of accessibility. That is why VIAA is supporting this broad sector initiative for media, archives and heritage.