Leobard's blog

automatt is happy about rawsugar. This flickr pic and the according text show why: rawsugar allows him to structure tags. As I predicted: folksonomies go towards semantic web.



Finally started to get into RawSugar. I imported all of my bookmarks from del.icio.us in about a minute, and now I can give my folksonomy some structure, and separate the tags I use for keeping track of stuff into different heirarchical categories or collections. And you can search your collections and your friends’ collections all at once!

We are going to present the current gnowsis alpha 0.8.x and the upcoming beta 0.9 at the cebit 2006. Hannover, 9. to 15th march.

I will be there from 9. to 12th March 2006. If you want to meet me, come during this time.

At the cebit: Halle 009, Stand B45

Features to expect there:

• from the epos project:
• context elicitation – what does a user do
• workflow support – what do I have to do now?
• miniqure, linker, lotsa features
• from the gnowsis beta:
• a complete new architecture, installer, components, database, … all is better in the beta
• a clear and easy way to write your own ontology to get your things together. Express your thoughts!
• desktop search as we always wanted
• tagging of e-mails, web-sites, files, ontology elements. all mixed together and really nice

as we are still working on it, I am eager to see what we have then 🙂

here is some official text from our news people:
The German Research Center for Artificial Intelligence, DFKI, is active in the field of innovative information technologies as Germany’s leading business-related research institute. It carries out research and development in the areas of knowledge management, intelligent visualization and simulation, deduction and multi-agent systems, language technology, intelligent user interfaces, image understanding and pattern recognition, robotics, safe and secure cognitive systems and information systems. An important element of DFKI’s mission is to move innovations as quickly as possible from the lab into the marketplace. Only by maintaining research projects at the forefront of science DFKI has the strength to meet its technology transfer goals.

so, after three years of doing a semantic desktop and facing our move towards the gnowsis beta 0.9 – what is the heart of gnowsis? Whats this all about?

In natural science – like physics or mathematics – people try to understand what nature is like. They can experiment how matter behaves and how the earth universe and everything works. Gods natural laws, they can find them out. In information science and informatics, we are creating means to express information and means to handle information. Information is human created – language is human created. It isn’t a natural law. So our work is more about psychology, philosophy and information science as such, and not about bits and bytes.

In the semantic desktop and our gnowsis implementation, I try to do what natural science tries for natural laws – I try to find a way to handle information that comes close the the “truth”. How do people remember information? How do we think? How do we represent concepts in our mind? Do we abstract from impressions and does our brain conceptualize, generalize? So how would we store information in a computer?

I try to hit this truth using cross-discipline thinking. Linguistics tells me that people conceptualize and that we have mental representations for sensory input. Philosophy tells me that each human being has a different view on the world (at least constructivistic view). Basically, each of us has his or her own weltbild. Our view of the world, our understanding what is truth and what is important, what happens right now, etc. But our sensory input is mostly the web at the moment – web tv, web music, web news, web-sites of your enemies (p.c. = competitors) and friends. And our output is not only spoken word but – hey – blogs, text, appointments, offers, contracts, etc. And the laptop on my lap is the way I interact.

So my laptop has to allow me storing my information the best way, a way that is near to the truth what information is like. Information is individually percepted, it is always understood differently, it means to me X when it means to you Y. The semantic desktop learns your Xs and Ys and tries to stick all the information you have as sensory input to these Xs and Ys. You can also enter your own concepts, subjects, topics, ideas, web-sites, whatever you encounter. The semantic desktop learns what you associate with your Xs and Ys. It suggests to you where to look for, it suggests to you where to put this document now, where it fits. It allows multiple places to put it. And “it” is your e-mails, your files, your web-sites, your photos, your instant messages. whatever is important for you.

Coming back to the heart – all these information bits are like sensory input – what we call data – and have to be cleaned a little to get to the level of information. Thats what we call at DFKI “leveraging”. Getting things together that mean the same for you. A person is a person, if it is on a photo, in an email or in an address-book, its always the same person. Gnowsis will go into this direction of connecting data bits from here and there to an information space for yourself. We published some papers about these ideas of ontologies and personal ontologies, where you express your mental model and link it to data.

I think that our approach is superior to others, because it tries to hit the truth behind personal information management. It ain’t just data, its data with meaning to people. I don’t know how long it will take to have a real semantic desktop and when we will be able to express our ideas on it, but I am sure that it will make life much easier when we can express and communicate ideas with it. I also expect that people’s use of computers will change when they can express their personal mental models better. Image school children of the future, that are taught by their teachers how to organize their thoughts. Teacher: “Today kids, we will learn the aspect of places. What is a place? Can you give me an example? Ah: School. Ah: the park around the corner where I can play basketball. Ok kids, go to your ‘places’ tab in the semantic desktop and add theses places there. Now each of you has two or three important places entered. Watch as the semantic desktop searches for public identifiers. See that tab showing with information about ‘Saint Franklins Elementary School’? Thats from wikipedia. So kids, now you will connect photos with these places. Have you got your phone-cams ready? … Ok, lesson ends. On my peer-to-peer net I see that some of you have annotated our school building very nice, and especially those web-graffity look wonderful. See you tomorrow.”

We should teach kids how to surf the web and how to remember stuff using computers. Isn’t that an important side-effect of writing something down: remembering it? We should teach how to build good folder structures and how to name files or how to publish something on the web so that it will be found when needed (perhaps years later).

Ok, that was carried away a little – the heart of gnowsis is not school. But I hope you got the picture – I am not content when I have all my Addressbook as RDF. I am happy when I have this, but not content. I am happy and content when the information lives in both worlds – data in my Address book and flickr account and information in my personal ontology that integrates it. Thats basically where we are aiming at and walking towards. A “memory extender”, not a bigger hard drive.

And by the way: we shouldn’t replace teachers, robots suck.

web desktop. A nice buzzword that fits into our semantic desktop thinking. Actually, the semantic desktop was called semantic web desktop on the first run.

Clearly the HTTP end-points on desktop applications would be useful to invoke an arbitrary set of methods on each application. I envision a system in which we can remotely manipulate desktop applications through HTTP. The same idea that I pitched a few years ago on top of CORBA.

thats something Sven Schwarz and me are talking about since 2004. So we have to talk more about it!

the best keyboard shortcut in eclipse is

ctrl + shift + l

it shows all keyboard shortcuts!