The Past, Present, and Future of Metadata
For many, metadata might be a topic that could be called, nicely, rather boring. For those of us who are jazzed by the concept of finding one cameo scene in the morass of all of the world's movies, or who want just to find the key points their professor mentioned in last week's lecture, advancements in metadata are quite welcome.
At the recent Open Video Conference, held in New York in mid-June, a session on time-based metadata was hosted by Devon Copley, Fellow at the Berkman Center for Internet and Society at Harvard University.
The round-table consisted of five participants, including: Matt Berry, CTO of online video platform Digitalsmiths; Chris Jackson, founder of the video distribution software organization MetaBroadcast; Andrea Rota, working on UTagIt, a media tagging service; Abram Stern, of the University of California Santa Cruz and MetaVid, the open video archive of the U.S. Congress; and Jörg Waitelonis, who has created Yovisto, an academic video search tool.
"We work with Tier 1 content owners and with most of the major studios," said Digitalsmiths' Berry, whose company was the sole commercial metadata application provider on the panel. "Our system does asset management, storage and playback, but the advanced metadata aspects are what sets us apart."
Berry showed off some of the features that are being integrated into his product.
"We've got the ability to cut the tracks of metadata much like the way an Avid would cut different tracks of video," said Berry. "We also have the ability to create custom metadata tracks, such as genre, since our studio clients have shown us that a genre may change from scene to scene within a single movie."
Trying To Make Metadata Portable
When asked about whether the metadata was tied to the clip, or as part of a separate database, Berry noted that the metadata is currently in a separate database, making it hard to make the metadata portable.
Other panelists agreed with him on the need to make metadata portable. Chris Jackson noted that URIPlay, an open-source project that MetaBroadcast has been working on with the BBC's backing, started as an interface project but has ended up as a metadata play.
"Our core goal is to help people find moving images," said Jackson, "but we've found we had to move well beyond user interfaces to the creation effective metadata tools and interfaces. We compared 'closed code' and 'open code' and realized that the pain threshold would be about the same initially for either type, but that the data range/quality of metadata could increase significantly if our open source code was used by the larger open source community. Our open software code revolves around metadata scraping and parsing."
Jackson's URIplay production version supports YouTube and a few key other vide platforms, with Hulu and others set to be released in a beta version, and an alpha version that will contain metadata parsers for the BBC, Channel 4 and a more general adapter set to be released some time in the future.
"More data is needed, and we're also heading toward caching and indexing," Jackson said, "where we can cache metadata that might be applicable across multiple clips (genre, actors, etc.) and are trying this out on a new site, amplus.tv, which was a one-week project leveraging all our previous learnings. Feel free to try out amplus.tv as it allows creation of playlists that 'curators' want to show off, with content from YouTube and other locations as well as an auto-generated RSS feed."
Waitelonis, a PhD student in Germany, has been working on the creation of an academic video search engine called Yovisto. Used for academic settings such as lecture recordings, seminars, or to record distance learning, Yovisto can catalog not only simple videos, but also those integrated with Powerpoint slides or web pages.
User interface optimized to let users drill down through customized time slices