Cluug at the SemTech2010, news from the Semantic Industry
What has changed about Semantic Technologies in the last year? Where are we now? Last week we were at SemTech2010 presenting Cluug and OrganiK and also learning what other companies are doing with Semantic Technologies.
Some things I (Leo Sauermann) noticed during the conference:
- We have really Cool UIs now based on Excel (a close phonetic match to Cool URIs which we had before already). This is a breakthrough. Why? Because the godfathers of Artificial Intelligence predicted that an Excel-Like interface is going to be the killer app, for example read this 2005 article by Jean Rohmer where he says exactly this (he was the first prolog guy in france, leading Bull's AI department for many years). Lee Feigenbaum's 5 minute pitch bout Anzo's Excel plugin was breathtaking - at least for me.
- It took 10 years of pitching by Tim Berners Lee and the SemWeb posse to convince people that putting data on the web is a good idea. It was worth it. Best Buy (story on readwriteweb) and Facebook (story on readwriteweb) are using RDFa now, which is considered a breakthrough.
- New Buzzwords appear now all the time, diluting the brand and standards already created. I hate new Buzzwords and broken Standards. I love the word "RDF" because it is a set of great standards that work great together: SPARQL, RDFS, RDFa, RDF/XML, TRIG - its all one data format. Great. The W3C in the Semantic Web Education and Outreach group did intentionally help Linking Open Data in marketing and publicity because it is well aligned with the standards and in general a great project (by people we all respect a lot, like Chris Bizer and Richard Cyganiak). I would not consider it random that Tim did shout "Raw Data Now" in his famous TED talk. What I hate is a word like "Open Graph Protocol" by Facebook. First of all, its not a protocol but a metadata markup. So the term is wrong. Second, they did break with RDFa best practices of other RDFa publishers. That is a high-nosed attitude. It means, the data you published already is useless for facebook and you have to markup everything twice - once for facebook, once for standard and community-compliant (like, dublin core) clients. If other companies follow, this is going to be a standards war. In their talk, they had the excuse that "this is easier for developers", but I oppose this. ATOM feeds are also not super-easy to do, and still people manage to write them. You just need to give them a template to copy-paste from. As facebook also gives a copy-paste template, they could have also given a best-practice compliant one.
- About buzzwords, I am neutral on "Personal Data Locker". Its a buzzword by David Siegel which he pushes in his book "Pull" (below a picture).
Ok, you need to buy David Siegel's book "Pull - The power of the semantic web to transform your business" ! Its the marketing brochure of the semantic web. It contains some examples and good ideas. To the right is a picture of me "pulling" a copy away from David Siegel. Ha!
I have a Pull on the book, it mentions our great project NEPOMUK on page 177! Sadly, David missed the point that this is the biggest and largest in-daily-use Semantic Web deployment so far. We (that is, Sebastian Trüg from Mandriva) integrated Virtuoso to the KDE desktop and there are now thousands of these neat triplestores on desktops, that is, alltogether, many more triples than in LOD. There is something missing - what he calls "personal ontology" is actually standardized as "Personal Information Model PIMO" also part of that operating system - and is also part of Cluug.com. So much for missing what has been around since 2006 and was deployed in a major operating system... but of course, our fault, we developers/scientists don't use our time for marketing but for delivering... which is probably because us Europeans have a different genome that USA people when it comes to marketing. I need an american in my company.
I love the word "Personal Data Locker" because it exactly reflects what I have invented in 2003 and been implementing in the Semantic Desktop standards and which we later have been rolling out at nepomuk.kde.org since 2008 to an audience of millions of KDE people. Technically, the only difference from the Social Semantic Desktop to David's idea is that we need a server counterpart, but that is what we are building with Cluug.com and some open source people also had that idea before (we called it the "Gnowsis server" all the time internally at DFKI). I hate the word "Personal Data Locker" because its a buzzword. I also hate the sentence "the personal data locker is a new concept by David Siegel". The right phrase in my stubbornly scientific ear would be "An E-Memory or Memex or Social Semantic Desktop is a concept that has been around since 1945 and the concept Personal Data Locker by David Siegel adapts it to contemporary people". In my eyes, there is no invention here. Of course, that is nitpicking by me who has been talking about it since 2002 and me being envy of not giving a keynote on SemTech :-) Of course, it may be that the word fades away like Memex/Xanadu did... and if you want to try out what it could mean, Cluug.com will be the first commercial personal data locker, once we are beyond alpha and beta stages.
I finally met Michael Bergman with whom I have been exchanging mails since sometime like 2006.
So in general I love the industry uptake of the Semantic Web, I personally hate the buzzwording of marketing. Well, to put it ironically there is some marketing pull-s*** available now which really helps us to explain the good working RDF and Semantic Web parts to the normal people from the street! Which is good. Buy the PULL book and give it away, dear readers.