Metadata & AI
Metadata Across the Supply Chain
Good metadata is a crucial component in any effective content supply chain. This means that interoperable metadata often sits at the heart of DPP technical work.
Metadata Exchange for News
When producing the news, time is of the essence. Good metadata, commonly understood between organisations, enables faster workflows. Read more here.
AS-11 file delivery
It's a core principle of DPP file delivery specifications that common formats for media are paired with common metadata as part of the delivery. Read more here.
Versioning with IMF
IMF offers powerful new ways to simplify the versioning and mastering of content. It provides many tools for carrying metadata, and we've released a DPP Recommendation on the carriage of metadata in IMF. Read more here.
Latest Metadata Updates
Artificial Intelligence (AI) and Machine Learning (ML)
The content supply chain is increasingly benefitting from automation and intelligence driven by AI and ML. The DPP has performed in-depth analysis to understand the maturity of these technologies, and is orchestrating a new initiative to bring together content and technology companies together to solve new problems and advance the state of the art. Read more here.
Latest AI Updates
Metadata Exchange for News
In a world of fast paced breaking news, efficient use of metadata is critical for news organisations. As global news outlets become more diverse, from traditional TV channels to online news networks, it’s clear that having the right metadata from the outset can help manage news production processes more effectively and get breaking news stories on air as quickly as possible.
The urgency of getting rushes back from the field to the newsroom means that, often, only basic metadata is captured (such as a date or location). This can add extra time to the production process as newsroom staff must add additional metadata retrospectively.
DPP Metadata Exchange for News is a new common metadata set for news assets, developed in partnership with news organisations and suppliers from the DPP membership. Use of this common metadata set promises easier search and discovery, clarity around rights information, and smoother collaboration.
Download The Documents
Sample files have been developed to assist with implementation. These are available to download below. The files match the 3 stages of the process:
These files provide an example of the metadata that is expected to be generated during news story planning. The metadata is expected to be transferred to acquisition/capture systems such as cameras, smartphones or other recording devices.
Acquire and Finish
These files provide an example of metadata that is expected to be generated during the recording and post production of news content. The metadata received from the planning stage is augmented with technical metadata generated as part of capturing the content. The content (media assets and metadata) may then be transferred to editing systems or to digital media asset management systems for further processing.
These files provide an example of metadata that is expected to be generated when the content is ready to be broadcast, published or distributed to third parties.
Latest News Updates
At the end of 2018, we published a Home Truths report titled AI for Real. It explored the maturity of the use of AI and ML across the content supply chain. We identified that AI was, in many cases, being developed in silos. It was sometimes a challenge for technology companies to develop highly accurate models due to the lack of accessible high quality training data, while content companies often lack the specialist skills to develop AI and ML solutions.
All of this resulted in AI being deployed for mostly generalist use cases, such as object detection and basic metadata automation. While the data generated can be useful, there are many issues still to be solved, including identification of unusual objects, understanding a wide range of accents and languages, and simply interpreting the volume of data generated by these generic models.
So the DPP set out to find ways to help media organisations solve some of these real remaining business problems through the use of AI. We'll be working in partnership with the Digital Catapult to run a challenge process that will bring together AI startups with innovative approaches, and content providers with practical problems to solve.
The challenge will be formally launched at a DPP Explore event in July, and further updates will be shared soon. To keep in touch with this work, and all DPP tech projects, make sure you're part of the DPP Tech Network.
To find out more or to get involved with this work, get in touch.
Programme Delivery Manager