Connected Broadcasting and the Metadata Challenge

Metadata is a big challenge for connected broadcasting. We need to find new ways to solve existing problems, whilst IP delivery has introduced a whole new set of challenges and opportunities.

IP delivery platforms have opened up the potential behind metadata...

IP delivery platforms have opened up the potential behind metadata...

Connected Broadcasting allows us to do a lot more with metadata. We can send more in-depth metadata than broadcast systems allow, and we can enrich that metadata from multiple sources. Connected Broadcasting combines lean-forward and lean-back forms of content discovery, enabling consumer-driven search and service provider-driven recommendations. Such use cases demand richer metadata.

The challenge is getting this metadata to flow through the delivery chain. Connected workflows are much more diverse than broadcast workflows, and everyone in the chain will have made different architectural choices. What kind of problems do we face?

  • Everyone stores slightly different information – for example, is a movie’s release year the global release year, or the year it was first released in each country?
  • While there were few drivers to interoperate metadata systems, organisations optimised their metadata around their business operations, rather than building their operations around a standardised format
  • Many organisations buy in their metadata and then enrich it for their own needs, resulting in a custom format
  • Connected Broadcasting frees us from the constraints of broadcast metadata, but this requires the aggregation of web and broadcast metadata formats
  • Company mergers often result in database integrations that are subsets of available metadata fields
  • Even where standards exist, many organisations have only implemented the parts of those standards that map to their current needs, making even metadata databases using the same specification not totally compatible

In addition, consumer devices implement different metadata formats. The result is that an end-to-end workflow in an ecosystem of content providers, broadcasters, aggregators and service providers will most likely need several metadata transformation and harmonisation steps, each of which introduces a risk of problems and errors:

  • The input data needs to be validated.  We can’t do this by hand, so we need automated approaches (for a technical audience: XML Schema Definitions, or XSDs) that can capture all possible metadata permutations.  But how do you validate the validation?
  • Where inbound data comes from multiple sources, each source needs harmonisation prior to being aggregated in a single database
  • Specific monitoring tools are needed to flag non-correctable data ingest errors

Each step in the workflow needs software to be written. It is challenging to find and fix all the inevitable bugs, particularly as we add support for more and more metadata formats, but if we don’t then errors will create problems throughout the workflow.

What’s the solution? In time, connected delivery will mature into a stable ecosystem like broadcast is today, with standards and testing regimes solving many of these problems. Until then, metadata management will remain a complex task that needs careful investment and time to mature. Arqiva has significant experience in building metadata management systems and knowledge of key industry standards. Our Connected Broadcaster and Connected Operator solutions feature proven metadata management and aggregation systems, which feature in key launches such as Freeview Connected.