A day in the life of LOD2 people…

… and life of the LOD2 in a day.

A plenary meeting for LOD2 project was taking place in Vienna this week (March 21-23). When Martin suggested we volunteer for the blog parade, I was the first to raise the hand, not because we were offered LOD2 mugs as reimbursement – I found this out after raising my hand — but to use this as a chance to show people what the “LOD2 project meeting” is all about, and after all… we at Zemanta are pro-blogging!

A lot has been said about the LOD2 project and the first day of the meeting, therefore I’ll skip the introduction and rather focus on integration, the main topic of the afternoon discussion on our 2nd day of LOD2 plenary.

Note: from this point on I’m assuming you’re familiar with the notion of Linked Open Data (LOD).

Dealing with open data requires lots of different kinds of (software) tools; first, data needs to be obtained one way or another, then cleaned, transformed to appropriate formats, so it can be enriched, linked… you name it. Follow me through this little exercise: imagine you are one of the poor souls… or let’s say open data pioneers, e.g. administrative workers, assigned to the task of making your agency’s data open, free, and re-usable. You may have LOD-aware friends who tell you about SILK or PoolParty, or they mention OntoWiki, LIMES, Sindice, maybe even LODGrefine(to name but the few).

Scream Cropped — Scream (Photo credit: Wikipedia)

At first you’re puzzled, but then you take the courage to google up the names. Wow! So many choices and functionalities, which ones to use? How to install this? What the heck is RDF schema?

You start dreaming how wonderful it would be to have all the tools in one place and someone to guide you through the whole process, letting you know what you’re supposed to do next…

Fortunately, your dream is a LOD2 reality (in making). The tools mentioned above are integrated into LOD2 Technology Stack and even more tools will be integrated into it. The best thing is the tools in LOD2 stack are freely available under Open source license. Sometimes semantic web dreams DO come true.

However, having all the tools in one place is only the beginning. What about the flow of information between components? We are currently working on making the workflow as smooth and user-friendly as possible. This includes integrated and unified user interface to provide shared user experience by exposing the functionalities of the tools that are important for the semantic web user.

It seems (business) people are more likely to trust data if they have some provenance information: who did what to which pieces of data from which origin. Sure, this requires authentication, which was also on our agenda for months now. Having a simple username and password in a local database doesn’t feel right for the world of Linked Open Data; authentication has to be webish. Besides, who wants to log-in into each and every one of the tools separately? Google, Amazon, Facebook, Twitter… they all made us sign-in-lazy by enabling the access to all sorts of services by 1-2 clicks. Fairly easy, right? Not really. Tools in the stack support different kinds of authentication (if any), some of them define user roles to grant access to certain functionalities. Should we use OAuth or OpenID? What about WebID? We had several (online) integration meetings – before this one in Vienna – to discuss pros and cons. Finally we decided to go with WebID and see what happens. If not sure what to do: 1. research, 2. experiment, 3. evaluate. Rinse and repeat.

And now something completely different. Another topic on our agenda for this day was also the development of application scenarios and testing of LOD2 Stack configurator. In my opinion one of the key elements in demonstrating the value of any tool are real life use case scenarios. LOD2 Stack is no exception. The Mihajlo Pupin Institute presented scenarios they’re working on, e.g. introducing LOD into their national statistical office (making it more transparent and accessible with LOD), airport emergency training and leveraging HRM corporate intelligence. We also heard a few words about LOD2 Management suite… and then it was already time to go to the Open Government Data Business Day 2012 – a call for action for both government agencies and SMEs to start contributing to LOD.
Open Government Data (OGD) BusinessDay event deserves a blog post of its own, but I’ll try to summarize it in a few points:

Power to the user: Vienna is empowering end-users by giving them means to check if they are really getting what they paid for (e.g. speed, latency of network connections).
Let them play: Open up data, let people look, take and play. We can all benefit from this.
Follow best practices: UK did it (data.gov.uk), so did Netherlands, Austria is doing it, what are you/we waiting for?
Gravity kick: If you open a little bit of data, it is hard to stop opening it.
Making data free pays: Change/update outdated business models, find new ways of getting the revenue. Don’t charge for data, make money from good services based on open data instead!
Killer app myth: Don’t wait for killer app to show you the $$ value of open data, it might be already there, but you just don’t see it. Remember email and SMS.
Barrier nutcracker: There will always be barriers on different levels preventing opening up the data. Brake them by building the community, open up discussion, make a dialog. And yes, DO send an email to government agencies if you can get the data you’re interested in. Guess what, you just might get it!
Dinosaurs beware: If you don’t want to die, keep up the pace. Walters Kluwer Deutschland – one of the players in the publishing industry knows this. They’re making their way towards becoming publishers of Open Data. Hell yeah!
Game of less & more: We need less data fragmentation and more data roaming. Elvis would say: A little less conversation, a little more action please.