In this very special episode of the DataPortability: In-Motion Podcast, Trent’s brother R. Mark Adams joins the data portability discussion. He is a genetic engineer who earned his Ph.D. in cell biology and was a pioneer in the field of bioinformatics. He is currently a Senior Associate at Booz Allen Hamilton and runs their bioinformatics group. Of specific interest related to data portability is his work for the open CaBIG (Cancer Biomedical Informatics Grid) project, a National Cancer Institute initiative to link cancer researchers and their data.
Up until now, we have focused primarily on the use cases around existing social networking websites. There is, however, a wealth of knowledge and experience to be tapped within other fields. Mark has worked for over 15 years designing and building large-scale informatics systems. Further, his extensive experience within the standards and open source communities place him in a unique position to provide valuable insight into issues being explored by the DataPortability Project.
During the conversation, Mark offered up some cautionary comments regarding the process of defining standards:
There’s a tendency on the part of industry, broadly, to try to skip to a technology stack as a means of adopting standards quickly.
One has to be careful in how one creates standards. This is why I say trying to divorce standards as cleanly as possible from their underlying technology implementations is important to do. The reason being it allows you to determine standards that can be widely adopted and used without the complexity or the risk of lock-in.
Rounding out the discussion was a call to action on both sides. Mark is reaching out to the DataPortability Project to become more involved in the bioinformatics field, and suggests we solicit participation from within their ranks.
The show is kicked off with a discussion about his recent speculation that Microsoft could buy Facebook and keep it closed. Scoble talks about the services and tools like FriendFeed that offer alternate news streams to counter the Facebook hegemony. The discussion also flowed around automated behavior tracking, advertizing, and the interplay between control/privacy within various portable data models.
Of particular interest is Scoble’s view of the inevitability of an open flow of user data:
Openness does win in the end. It will just take a little bit of time to get there. We’ll see a lot of new stuff come along to make it easier for users to open these systems up.
For Plaxo, we have Joseph Smarr, Chief Platform Architect, and John McCrae, VP Marketing, talking about the acquisition and how it furthers data portability. Specifically, Smarr made it clear that the name of the game in portability is not making everything homogeneous, but rather opening up the flow of communication across systems:
Data portability is about empowering users to connect the tools they use so they don’t have to repeat themselves over and over again. So that the information can flow for others to discover it. It would be a mistake to characterize it as making everything exactly the same.
On the same thread, returning guest Kevin Marks, Developer Advocate for Google’s OpenSocial project, highlights their commitment to the openness of data portability:
One company can’t hold anything hostage because we’re connecting together open standards. All these pieces can be supplied by multiple parties. You can interoperate without having to have a business negotiation because you can write to the standard and the standard works.
In the discussion, Marks also corrects some common misconceptions around Google’s Friend Connect. Some of the reporting about it mistakenly assumed that Google would be siphoning off the friendship graph when using it’s system to connect sites. He clarifies that Friend Connect enables the portability of user data by mapping the connections, and isn’t storing the data itself.
There is a lot of focus in the DataPortability Project about making it easier to access user data. Another aspect to data portability, in general, is an analogous set of activities around enabling other data on the web to be more machine accessible. A few groups have been approaching this issue in various ways, many of whom work under the umbrella of the Semantic Web community. One subset of people focusing their efforts on this are taking what they call the Linked Data approach.
At a recent Cambridge SemWeb Gathering at MIT, Kingsley Idehen, CEO of OpenLink Software and a founder of the DBPedia project, had a great term for where he sees himself within the greater context of people working on these issues:
I like to say that I belong to the Semantic Web Community, but I’m a member of the Linked Data Tribe.
I found this concept of a tiered relationship and allegiance illuminating. Talking about it with him, he makes a distinction between the community as a whole and the fact that he focuses on a specific set of actionable efforts. It has been this sense of “what can be done right now” that has helped build upon what others are doing to move toward the goals of the community as a whole.
For example, I recently discussed how microformat markup could benefit the Semantic Web with Danny Ayers, an RDF/SemWeb guru working for Talis. Similarly, Ivan Herman gave a talk at the gathering about how to leverage RDFa within the context of an existing XHTML web page. Both examples are stepping stones in the direction of truly portable data on the web, and something that Kingsley considers the “data substrate” upon which Linked Data representations can be built.
To that end, I’m on a mini crusade to encourage developers to take the extra few minutes required to consider how their display layers can expose their content with effective markup. Rather than everyone having to learn OWL, RDF, and SPARQL before any progress can be made, there are some simple steps that will catalyze further steps. It’s really not that hard, and even if you’re not a developer you can mark up your own blogs and pages with microformats to provide search engines with much-needed context to describe your content.
NOTE:I’m purposefully not diving too deep here into the real “meat” of Linked Data. Instead, I hope you’ll spend a couple clicks checking out the simplicity of what can be done to help build the “data substrate”.
We’re also joined by Brady Brim-Deforest, founder of Human Global Media, talking about the DataPortability Legal Entity Taskforce. He provides a good overview and update on the process underway to formalize the the project under a recognized legal banner.
The featured interview segment is with Danny Ayers, Semantic Web Developer at Talis. He touches on moving from document linking, through microformats, to feature-rich RDF modeling to identify portable data. Contrary to popular belief, he dispels the myth that it’s hard to migrate from a standard SQL data representation into addressable semantic objects.
