Valentin's Blog

To me Electronic Health Records (EHR) are one of the most exciting IT-developments of our age: the potential for better medical care, creation of new knowledge through mining this data and the potential cost savings are just so extremely large. And with the American government having dedicated $19 billion to speed up the computerization of medicine, we may actually see some real improvement in the next few years.
Two articles in the New England Journal of Medicine now give interesting perspectives on this topic.
The first article shows just how far away the vision of a comprehensive EHR system still is:

[...] only 1.5% of U.S. hospitals have a comprehensive electronic-records system (i.e., present in all clinical units), and an additional 7.6% have a basic system (i.e., present in at least one clinical unit). Computerized provider-order entry for medications has been implemented in only 17% of hospitals. [...]

The second article is more of a position statement that details very clearly what we should have learned from the web about the design of an EHR platform:

The platform approach to software design can be used to create and sustain an extensible ecosystem of applications and to stimulate a market for competition on value and price.

First, there should be liquidity of data. The platform and its applications should reduce impediments to the transfer of data, in an agreed-upon form, from one system to another.

Second, there should be substitutability of applications. The system should be sufficiently modular and interoperable so that a primary care provider could readily use a billing system from one vendor, a prescription-writing program from another, and a laboratory information system from yet another.

Third, the platform should be built to open standards, accommodating both open-source and closed-source software

Three points that should really be the blueprint for all large government organized IT-projects! (and, in fact, all large enterprise IT projects as well)

My presentation from the STI Industry Day - the first such networking event organized by the German branch of STI International (the organization of scientific, industrial and government parties interested in fostering semantic technologies)

The presentation addresses the curiously still unsolved question of an understandable definition for "semantic technologies".

Was sind eigentlich semantische Technologien

View more presentations from vzach.

"The Big Picture" of the Boston Globe is just one of the most amazing websites around. Every few days they post a set of the most gorgeous, intriguing and thought provoking high resolution pictures about some topic, like Carnival, Kyrgyzstan and the American air base there or - and that brings us to the title of this post - Robots. Go on, have a look, you won't regret it. 

One of the most exciting developments in IT these days is that computers are starting to see, are starting to become able to understand and reason about pictures and movies. However, this development also leads to a host of new privacy nightmares.

I'm thinking about this development for quite a while (see also this post on "Smart Spy Cams" from January 2006) but even I was  shocked by some of the applications that smart spy cams are finding (from a very interesting Economist article):

Soon the company will start selling a 'smile measurement' system that will alert managers-in real time, if desired-when a cashier fails to muster an adequate grin. The software is configurable, so employers will be able to decide just how happy their employees should appear.

You should read the entire article for more useful and perverse applications of machine vision technology.

There has been quite a bit of discussion about Tim's appearance at TED - so, here's the video:



Obviously not a technical talk; nevertheless I was quite amused that neither OWL nor Ontologies were mentioned at all :)

These days I'm more and more amazed at how the increasing alarming global warming rhetoric is not matched by a willingness to do the necessary thing to actually tackle the problem. And I'm not even talking about the difficult political questions of how massively lower our energy consumption - here I mean the research that we need to cope with Global Warming that is not funded because of ignorance or ideology. Let me elaborate the three most striking examples:

Nuclear Energy (from Thorium): Humankind's unprecedented wealth largely hinges on the availability of very cheap energy and - consequently - it seems tremendously naive to assume that a large scale revamping of our entire energy infrastructure around often volatile and more expensive renewable energy sources will be possible without making us all a lot poorer (shaking the very foundations of our societies in the process). For this reason it seems awfully negligent to pursue nuclear energy only half heartedly or (as in Germany) not at all - afterall nuclear energy seems to be the one CO2 free energy source that requires the least adoption of our entire infrastructure (except CO2 storage, perhaps). Particular puzzling is the little investment and attention that nuclear energy from Thorium is seeing, since even though it would require a very large initial investment, it offers a chance for nuclear energy without the risk or a catastrophic meltdown, the prolifiration of nuclear weapons, the problem of decreasing Uranium reserves and nuclear waste thats toxic for millions of years (almost all trash from Thorium reactors would decay within a few hundred years).   Its really a technology that could be a tremendous help for the years to come - but you don't hear about it. You can learn more about energy from Thorium from the TechTalk below:


Geoengineering: Many ambitious plans for curbing greenhouse gas emissions only slightly slow down the increase in global CO2 concentrations - and even these plans are enacted only very slowly or not at all. And even a radical cut in CO2 emission will not lower CO2 concentrations - these will only come down over a long period of time. In short: cutting emission (while indispensable) offers no way to quickly combat Global Warming or its effects. Now this is a big problem: we are already seeing the ice caps melting faster than expected and it may yet turn out that Global Warming will cause a global emergency of tremendous proportions (even while we work to cut emissions). Or - even worse - that there really is a climate tipping point where increasing global temperature becomes a self-reinforcing process.  And again: cutting emissions offers no way to quickly do anything about it. However, there are really good arguments that geoengineering (such as sulfur seeding in the stratosphere) could quickly, cheaply and effectively lower the temperature in such an emergency. There are also many problems with this specific proposal and geoengineering in general - but it seems downright stupid to not at least massively invest in exploring and understanding these methods and their drawbacks! Below I've embedded a great TED talk on this topic - should you be interested in learning more.


Global Carbon Accounting: Amazingly there is no real global network to measure CO2 emissions and - as a direct consequence - our carbon lifecycle models cannot explain our observations: roughly 20% of the CO2 created by humans is simply not accounted for, we simply don't know where its going. Now there have been attempts to somewhat close this gap by sending a carbon measuring satellite into the orbit but sadly, it just failed to reach the orbit. Luckily the Japanese Ibuki satellite launched at the end of January may also fill this role. These measurements from space are not nearly as precise as those using sensor stations at the ground but at least - for the first time - we may get a really global picture.  There is a lot more on this problem in the TechTalk embedded below (the measuring problem is discussed explicitly around minute 15; the satellite he's talking about is the one that crashed):