M+E Daily

GiantSteps’s Rosenblatt: Rights Metadata Standards Important, But Not All are Created Equal

NEW YORK — Creating rights metadata standards are important for industries that use such information, but not all standards have been successful to date because some of them haven’t been well thought-out and created for the correct reasons, according to Bill Rosenblatt, president of GiantSteps Media Technology Strategies.

Rosenblatt has played a role in the creation of a few rights standards initiatives, starting with the Digital Object Identifier in the 1990s and the MPEG Rights Expression Language, he told the Rights & Metadata Madness conference July 20. Despite that, “I consider myself a standards skeptic,” he said.

“A lot of standards fail,” he pointed out, going on to provide what he called his “standard handy, dandy test for the likelihood of success of a standards initiative.” His criteria for a good standard are based on the correct answers to several questions, he said: “Is the scope focused and clear. or are we trying to boil the ocean or solve world hunger? Are we trying to do something relatively simple, or are we over-engineering? Or is it a camel that we’re building, which everyone knows what a camel is: It’s a horse designed by a committee.”

“Redundancy” is something that must be avoided at all costs before creating a standard, he said. “Are we innovating, or are we building on existing solutions that are in the market? Or are we trying to reinvent the wheel? Are we building something that’s a win-win-win for everyone involved, or are we building win-lose kind of scenarios? Are we building something that’s going to work fairly well with companies’ existing processes and systems, or are we going to build something that requires replacing and gutting what we have in place?”

“And then, finally, are we trying to solve a known, practical problem – whether that’s a business problem or some other type of practical problem? Or are we just creating a standard so that we can all be nice and get along?”

There are several components to a rights standard, he went on to say. First, there are “controlled vocabularies” – a list of terms or words that are allowable for a certain metadata field, such as a type of right being granted, he said. Then there are the data formats. Rights expression languages, meanwhile, are the “grammars that allow you to put together expressions of rights-related information that involve actions in relation to content and entities such as licensors, licensees, etcetera,” he said. “And then there is the concept of a registry, which is a database accessible online of rights information.”

One important thing that must be decided by organizations looking to create a rights standard is what they want to do with it, he said. “You could standardize on data formats in rights management systems from vendors like some of the ones who are sponsoring today’s event.

You can use them for communication downstream to your licensors or other entities involved in distribution of your content for humans to look at for non-automated processing purposes. And a lot of the activity has been there because people are, in some ways, reluctant to go full-on into automation,” he said.

“Standards are also useful for rights enforcement – if you’re using DRM, for example, or if you’re using a copyright monitoring scheme of some sort. You can use them for integration internally between your digital asset management and your contract management systems.” A standard can also be used for the “look-up and discovery of rights information,” he said, giving as an example: “Here’s some content that I’m thinking of using. What rights are available and under what terms?” And then there is the “holy grail” — B2B licensing automation, “where you can actually automate these things,” he said.

Rosenblatt went on to cite several standards, starting with the Open Digital Rights Language (ODRL) — a rights expression language that was invented in 2000 and served, for a while anyway, as an “ad hoc rights standard” before recently becoming a World Wide Web Consortium (W3C) standard that “competed” with a standard that Rosenblatt said he was involved with: the MPEG Rights Expression Language (REL). ODRL has wound up “gaining a lot more acceptance than MPEG REL did” and has become “the basis for a growing number of standards, particularly in the license automation arena,” he said.

In the news industry, there’s been RightsML, created by the International Press Telecommunications Council (IPTC), and consisting of a rights expression language, a data format and controlled vocabularies for that data format, he said. It’s focused on news articles and other news content, including photographs, and was designed originally as a “replacement for a standard that failed” called the Automatic Content Access Protocol (ACAP), he said.

The effort was designed to automatically send a message to Web search engines including Google something along the lines of “We would like it if you didn’t crawl this content and make it available in your search results for X number of days or hours,” or “We would think it would be great if you could please, please, please only show this photo as a thumbnail in your search results,” he said. The problem was Google refused to go along with it, “so that standard didn’t work, and RightsML has since kind of migrated to other use cases” as a result, he said. “But it’s a very interesting standard” that he said focuses on downstream licensing.

There is also the PRISM Usage Rights standard from Idealliance, a metadata standards initiative that grew out of the consumer magazine industry, he said. That’s basically a data format with controlled vocabularies that was based on the Extensible Markup Language (XML), he said. There’s another PRISM-related rights standard in the magazine industry called PRISM Rights Summary, which was designed for the integration of an organization’s digital asset management system with its contract system, he said, adding it’s a more recent standard that was based on ODRL.

“In the photography sector, there’s the Picture Licensing Universal System, an ad hoc standard that he said was “gaining some acceptance.” It’s used if you are a producer or art director at an ad agency and want to license photographs to be used in a Chevy ad and you want to do that so the rights processes are automated, or if you are a photographer and want to license your work to licensees, he said. The open standard was based on XML RDF from the Adobe XMP world, and has been “adopted very strongly” among Adobe users, he said. The focus there is on downstream licensing, he said.

New initiatives worth looking at include the W3C Permissions and Obligations Expression Group, which is creating a standard around Web content, including Web pages and things that can be put on Web pages, he said. It’s not about protecting content, but rather about advising anyone who wants to use content what permissions its rights holder might be offering and what obligations they might have if they use that content, he said. “That’s how they’re positioning it. They’re just getting started. So it’ll be interesting to see what happens there,” he said.

In the wake of the Google book publishing settlement, the Book Industry Study Group (BISG) created a controlled vocabulary for book rights information, but then the settlement was rejected by the judge in that case, he said. As a result, BISG “kind of shelved” that effort he said. But the rights standard study group was recently “revived” and Rosenblatt is part of the group that is studying what the business case would be for such a standard and what it would be used for, he said. “There are answers, of course, but we haven’t converged on what those answers are for the purposes of standards” yet, he said.

In the music industry, rights discovery is a “big, big problem” that’s “unsolved,” he also said. If you want to offer music, for example, as part of a mash-up or a service, you would want to know who owns it and what rights they have, he said. There is a need for a rights registry because, although there are a bunch of small or proprietary ones, there is no big, global one yet, he said. “There have been various attempts to solve that problem, the most high-profile of which” was the Global Repertoire Database that started in Europe, but “collapsed a couple of years ago” and is now “essentially dead,” he said.

The Open Music Initiative is a new effort that he’s involved with that arose out of a conference at the Berkeley College of Music in Boston, he said. “It’s very, very early days, so we’re not exactly sure what’s going to come out of it yet,” he conceded.

“There are a lot of challenges to the idea of adopting standards for rights,” he said, pointing out they “take time to achieve momentum.” There are also “a lot of companies that do need to get their internal houses in order when it comes to rights information,” he said. Although organizations and companies such as the New York Public Library and Viacom have done a “great” job, many others have not, he said.