This video is probably worth it only for Tim O’reilly and his discussion of what Web 2.0 is:

I read a paper on the Google Similarity Distance this morning by Cilibrasi and Vitanyi. They search for word cooccurrences using the Google search engine. Their formula goes as follows: (G(x,y)-min(G(x,x),G(y,y)))/max(G(x,x),G(y,y)) where G is the “Google code” function. The Google code function is defined as -log g(x,y) where g(x,y) is the normalized number of web pages containing both term x and term y: the normalization is such that if you sum up g(x,y) over all x,y then you get 1.0. With this simple approach, they seem to be able to translate between English and Spanish, build a thesaurus, and so on. This reminds me a bit of the recent work done by Turney on analogies.

The Australasian Database Conference (ADC2008) will be held at the University of Wollongong. It looks like a pretty good generic database conference.

In the last year, I decided to go back to science-fiction novels and learn what is out there. Here are my favorite scifi authors.

• Frank Herbert for his Dune saga. Some interesting trivia: Frank is a college dropout and Dune was rejected by nearly twenty publishers before being accepted by a small publisher (Chilton). Among other things, Star Wars is a Dune derivative. Among ideas explored are: AI is dangerous and should be forbidden, seeing the future can be a curse, human beings should think long term. He became a full time writer only at the age of 52. His son Brian tried to expand on Dune, but do not bother: he does not get Dune.
• David Brin for the Uplift saga. David has a Ph.D. in space science. His work is concerned with ecology: how do species and intelligence evolve and how can we protect planets? I find it fascinating that we may, one day soon, allow monkeys to speak (why not) through genetics.
• Dan Simmons for Hyperion. Oddly enough, Dan writes about AI and its consequences for humanity, but it is not immediately apparent when you start reading him. Very original. Are you a slave to your technological devices? How do you know you are not being fooled?
• Georges-Jean Arnaud for his Compagnie des glaces saga. The saga is made of over 60 novels! He describes life after a catastrophe that has left Earth frozen. In order to survive, the few surviving human beings become dependent on sophisticated railroad companies. Other than the terrible climate and sex, the main theme is the importance of anarchists. Unfortunately, Arnaud is probably the greatest French scifi writer of the XX^th century: his work is deep and ambitious, but incredibly messy. He has other good scifi novels.

I plan to try my luck next with Iain Banks. I am actively looking for recommendations on what to read next.

I did a little bit of research on the Web today, following a statement by my friend Owen to the effect that yield returns were not something Joe Programmer could be expected to know about.

Turns out that yield returns are supported in C#, Visual Basic (VB.NET), Python and Ruby.

So, yes, I’d say that if you are programmer out there, you ought to know about yield returns.

The New Scientist (and slashdot) reports that a very simple method that basically achieves what million-dollars Quantum Cryptography set out to achieve: unbreakable two-way communication. This is due to Laszlo Kish. His papers are on arxiv and they appear pretty convincing, but I have left the world of Physics a long time ago.

Bruce Schneier describes it in those terms:

How would you feel if you invested millions of dollars in quantum cryptography, and then learned that you could do the same thing with a few 25-cent Radio Shack components?

And he concludes:

Basically, if Kish’s scheme is secure, it’s superior to quantum communications in every respect: price, maintenance, speed, vibration, thermal resistance and so on.

Is this true? Is Quantum Cryptography obselete?

There seems to be theoretical difficulties with Kish’s approach, but his experiments seem to suggest that he got it right.

The New Scientist (and slashdot) reports that a very simple method that basically achieves what million dollars Quantum Cryptography set out to achieve: unbreakable two-way communication due to Laszlo Kish.

Bruce Schneier describes it in those terms:

How would you feel if you invested millions of dollars in quantum cryptography, and then learned that you could do the same thing with a few 25-cent Radio Shack components?

And he concludes:

Basically, if Kish’s scheme is secure, it’s superior to quantum communications in every respect: price, maintenance, speed, vibration, thermal resistance and so on.

My current employer, a state university, may technically go out of business in July. Naturally, management blames the lack of funding. There are two issues at stake.

Firstly, the University has been chronically in the red. This means that tuitions and government grants are not sufficient to cover cost. Meanwhile, the number of faculty members has gone up 5% in the last few years. So, surely, someone expected the University to have increased revenues. But from where?

Secondly, the University has invested massively in real estate. I mean millions and millions of dollars in buildings. What for? To accommodate the increasing number of students? Actually, the number of students attending classrooms has been pretty stable in recent years. Whereas the business students keep on coming, the sciences are not attracting more students as time passes. Distance education is another matter: the number of students taking online courses is growing by more than 10% a year. But as anyone knows, distance education does not require large buildings.

I find this interesting. Isn’t it the job of managers to do the best they can with whatever budget they have? If you are a highly paid university manager and you spend and spend without counting, then where is the challenge in this?

It seems that spending the money then turning around and say “oops! we spent money we didn’t have” is irresponsible. You’d think people would get fired over this, but they won’t.

What is going to happen? Clearly, less popular courses and programs will have to go. There is no other way to reliably save money when you are a university. The only recipe for profit is to focus on the profitable programs (Business!!! more Business!!!) and to slash less profitables one (Physics?). Except for distance education, of course, which will continue to slowly grow, unnoticed, and without any expensive building to support it.

(No, I do not expect to lose my job. Not for now anyhow.)

Though blogpulse seems to be going nowhere, as far as I can see, it is still one of the most fascinating tool out there. What it does is plot word occurrences versus time on the blogosphere. The recall is rather poor compared to Technorati but the time series plot are very nice.

Here’s one comparative plot that a student in my Information Retrieval course (Mahmoud El-Bachir) has submitted:

You can see clearly when Christmas is (Noël in French) and when the new year is… I think you also have the Chinese New Year too! (Seek the smaller bump).

My only beef is that I do not have access to the raw data: it would be really cool to build applications on top of blogpulse, but I guess it goes against their business model.

Technorati is one of the few long standing Internet companies to be in the search business and to stand tall in front of Google and Yahoo. And there are only 45 of them. That is not exactly a basement operation, but given how well established the company is, that is a very small number.

What does it say about the future of Information Technology? Are all the mail servers in the world be run by 30 people one day?