PubForge Blog

What have we done with the new NPR API? This would be a good place for people to share examples of cool mashups and apps they’ve devised to tap NPR’s open content. Or to suggest ideas on which we could perhaps collaborate.

Here’s one of mine: What if I tagged my news stories, interviews, and other content with keywords based on the NPR taxonomy (or even just my own keywords) and when I publish that content on my website, it generates a query to the NPR API? I could have a widget sitting next to my content pulling in related stuff dynamically. As the NPR API expands its reach to other public media sources, each content entry on my site becomes an entry point to a growing universe of related content. Of course that universe might get pretty big, so what if we wrote a script that could then parse everything and generate a navigational structure based on the metadata returned with the results of the query. So the query results would evolve over time, and so would the navigational structure.

What if next we expand the range of sources to query based on the metadata of our initial content? Scientific and cultural institutions have large collections of content, many accessible through an API. Funding is increasingly premised on open collections and public research results. What if we tap into that, so a given media object can serve semantically and programatically as merely a starting point to explore a growing web of deep knowledge and perspective?

Maybe that kind of language sounds cheesy or something, but it seems like a fun thing to me.  On the other hand, maybe let’s start with the NPR API and go from there…

Jack Brighton
WILL Public Media

Those of us consumed with passion about metadata for A/V objects (and who isn’t…) have been excited by the emergence of the PBCore. We present here an update.

In our last dramatic PBCore episode, CPB funded a multi-year project to develop a standard for shareable metadata about audio and video productions and files. This culminated in the release of the PBCore Data Dictionary and an associated XML schema, with Version 1.0 in April 2005, and an improved Version 1.1 in January 2007.

We’ll leave an actual description of PBCore for another time and place, or get full details on the PBCore site.

It turns out PBCore is darn useful. Film archives, academic media collections, and media curators including the Library of Congress are actively pursuing systems that speak PBCore. Not to mention PBS, NPR, and a growing number of local stations. At recent AMIA conferences PBCore has been a central topic. PBCore has become relevant and possibly important to all moving image archivists, because it fills a black hole in the metadata universe concerning digital media.

Color us surprised when the initial CPB project to develop and support PBCore ran out of money last August. Forthwith, the principle developers at WGBH and elsewhere proposed a second phase, to establish a PBCore change-management process, plus funding to maintain the website, workshops, and other support activities. So far the response from CPB has been PBCore who?

What’s at stake? Considerable time and intellectual effort to develop a really good standard for A/V metadata, something everyone in the moving image community needs. Plus a certain (large) degree of credibility, because CPB was leading the PBCore effort and now we’re in some danger of abandoning PBCore. Like we somehow just forgot about it. With this project, Public Broadcasting has been a hero to the librarians and archivists, but it looks like we’re dropping the ball just when everyone wants to play.

So let’s some of us carry the PBCore torch for the next bit while pushing for further CPB action. We might have to get a bit militant. When someone ticks off the librarians, you don’t want to see what happens. Or maybe you do…

Jack Brighton
WILL Public Media