The Software and Data Licensing Solution: Not Your Dad’s UBMTA

Editor's Notes

• #39: Abstract: Biomedical data integrators grapple with a fundamental blocker in research today: licensing for data use and redistribution. Complex licensing and data reuse restrictions hinder most publicly-funded, seemingly “open” biomedical data from being put to its full potential. Such issues include missing licenses, non-standard licenses, and restrictive provisions. The sheer diversity of licenses are particularly thorny for those that aim to redistribute data. Redistributors are often required to contact each sub-source to obtain permissions, and this is complicated by the fact that on each side of the agreement there may be multiple legal entities involved and some sub-sources may themselve already be aggregating data from other sub-sources. Furthermore, interpreting legal compliance with source data licensing and use agreements is complicated, as data is often manipulated, shared, and redistributed by many types of research groups and users in various and subtle ways. Here, we debut a new effort, the (Re)usable Data Project, where we have created a five-part rubric to evaluate biomedical data sources and their licensing information to determine the degree to which unnegotiated and unrestricted reuse and redistribution are provided. We have tested the (Re)usable Data rubric against various biomedical data sources, ranking each source on a scale of zero to five stars, and have found that approximately half of the resources rank poorly, getting 2.5 stars or less. Our goal is to help biomedical informaticians and other users navigate the plethora of issues in reusing and redistributing biomedical data. The (Re)usable Data project aims to promote standardization and ease of reuse licensing practices by dat providers.   
• #46: Clearly stated A clearly stated, unambiguous, and hopefully standard, license for data use is critical for any (re)use of data: if there is no license to be found, then rights are unclear and one needs to assume the default: all rights reserved. more » Comprehensive and non-negotiated Data that is mixed under different licenses, only partially available, or must be in some way negotiated creates barriers to the (re)use of data. more » Accessible Data must be accessible in a reasonable and manner to be useful to the broader community. more » Avoid restrictions on kinds of (re)use Data should be able to be copied, built upon, edited, and modified as freely as possible. more » Avoid restrictions who may (re)use Data should should be available to as many people as possible for their (re)use.
• #52: Overview, how many at which scores
• #57: TTOs are not able to address all four of the above bullets. There are no off the shelf noncommercial licenses that adequately protect our royalty streams, allow for rapid data aggregation between academic institutions, and safeguard the provenance of research data. In a broad sense, we TTOs are undergoing a coordination breakdown.
• #58: The root cause of the coordination breakdown is that there are too many rightsholders. It is subtle, but the end result is massive underuse of data. Commercial R & D that is less efficient and effective. Same goes for university R & D and for clinical informatics. A coordination breakdown like this is called an anticommons problem. But to unpack how an anticommons works, let’s start with the commons problem.
• #59: In 1992, Canada declared a moratorium on the Northern Cod fishery, which for the preceding 500 years had largely shaped the lives and communities of Canada's eastern coast. It has been over 20 years since the moratorium on fishing Atlantic cod in eastern Canada, but the fish stocks have not replenished. Or in other words, there are no fish. Commercial overfishing is why there are Somali pirates on the high seas. They used to be fisherman. When we use the term Tragedy, we are not being dramatic.
• #60: The Rhine River during the Middle Ages is the best example of a Tragedy of the Anticommons. The Rhine starts in the Swiss Alps, follows the Franco-German border, flows through the Netherlands and empties into the North Sea. It was an important waterway for commerce and Europe’s principal highway during the Middle Ages. The Roman Emperor protected the great European trade route. Merchant ships paid a modest toll to safeguard their transit. But after the empire weakened in the thirteenth century, German Robber Barrons built castles on the Rhine and began collecting their own illegal tolls.
• #61: Robber Barons built spectacular castles along the Rhine, including this one. There were hundreds of castles, including 35 within a 90 mile span. Far too many. The gauntlet of tollbooths made shipping impractical and commerce down the river halted. Today, the hundreds of ruined castles are a lovely tourist destination…But for hundreds of years, everyone suffered—even the barons. The European economic pie shrank. Wealth disappeared. Too many tolls meant too little trade. This is why I say we Universities are medieval.
• #62: As tech transfer professionals, we should be most fearful of gridlock, because it means rampant underuse. In the data sharing world, we have underuse due to high transaction costs.
• #63: The above continuum was introduced by Columbia Prof. Michael Heller, a thought leader in this field. He says there’s no effective ownership at the ends of the spectrum… Overfish the cod, and there’s effectively no cod to fish anymore. Too many data license restrictions, then data aggregation and integration is impractical and essentially impossible. So what’s the data licensing solution? Some say, nuke any and all IP rights in the data – essentially the copyrights in the databases. That’s called CC0. That involves waiving all interests in one’s works and thereby place them as completely as possible in the public domain. We’ll talk about that in a second, but first consider how governments deal with the commons problem. They can’t control the common resource like the ocean, but they can declare moratoria on how you access the ocean, i.e. how and where you fish. What it effectively does is move the resource closer to the middle of the continuum. I think that’s the real solution for data licensing. But first, let’s cover the off the shelf data licenses.
• #64: CC BY This license lets others distribute, remix, tweak, and build upon your work, even commercially, as long as they credit you for the original creation. CC BY-SA This license lets others remix, tweak, and build upon your work even for commercial purposes, as long as they credit you and license their new creations under the identical terms. This license is often compared to “copyleft” free and open source software licenses. CC BY-NC This license lets others remix, tweak, and build upon your work non-commercially, and although their new works must also acknowledge you and be non-commercial, they don’t have to license their derivative works on the same terms. CC BY-NC-SA This license lets others remix, tweak, and build upon your work non-commercially, as long as they credit you and license their new creations under the identical terms. CC0 owners of copyright- or database-protected content to waive those interests in their works and thereby place them as completely as possible in the public domain, so that others may freely build upon, enhance and reuse the works for any purposes without restriction under copyright or database law.
• #66: NIH Scientific Data Council nearly adopted the plan to declare NIH-funded research data as CC0.