Tumbling Towards Enlightenment

Actually I wanted to post a link to Flight Radar 24 already a while ago, but the current eruption of a icelandic volcano makes it even more interesting to look at the live air traffic situation, as the all planes remain on the ground in UK and whole of Scandinavia.

 The map features detailed information of each major airplane currently in the air, with complete path history, altitude, etc. So, next time you look up into the sky and wish to fly away as well, you can go to the site and will probably find out that the airplane either goes only to London or Frankfurt anyways :-)

There is also a similarly cool visualisation of the live Marine Traffic:

Additionally, the Flowing Data blog posted an overview of cool traffic visualizations just today. Check it out.

crazily playful, i just love it

(to put this in somewhat random business context, read this article on The Secret To Great Work Is Play)

A map of genetic variations and correlations in Europe

—> So, the Finns are indeed quite different :-)

Not that I’ve got any business hanging out on online dating platforms, but OkCupid.com seems to be quite determined on finding a perfect match for you and they use quite some number crunching to improve their matches. I am lucky to be in a great relationship, so that I don’t need to rely on some algorithm to find me a partner, as this just feels a bit unnatural to do. Recommended reading, recommended listening (which I am actually a fan of), recommended (sex) partner for life,… strange strange world. 

Anyhow, OkTrends is the official blog of OkCupid, and they publish super interesting statistics and findings on a regular base. This seems to be a good example of an online service which sees the data collection craze not as an end in itself, but rather to improve their user’s lifes and along that way also take some time aside to feed back valuable insights back to the global community.

Take these recent posts for example: The 4 Big Myths of Profile Pictures (with 870.000 Reads), Your Looks and Your Inbox, or The Case for an Older Woman (in my opinion the best written article so far). Besides providing single figures conveying strong revealing insights already (“2/3 of the messages sent by males go to the top 1/3 attractive women”), the blog authors heavily use visualizations (sometimes of the interactive kind) wisely to get their points across:

It helps if women display cleavage on their profile picture. It helps even more, the older they get:

Mature women that are dating online are more self-confident and happier with their lifes in general:

However, mature men rather prefer younger women, and quite often even go for the much much younger girls:

Quite revealing, I would say.

On a final remark on that topic: There is some great value in having bad dates, that sites like OkCupid are (probably?) not able to catch. These dates make it clear to you what you are really looking for, and are oftentimes also then the best stories to tell :-)

Yet another public source for information on a particular business/subject/whatever: Fan Page Analytics. It currently only analyzes Facebook Pages, but it seems that there are plans for Twitter and MySpace as well.

Here is a world map of the relative popularity of the NOKIA Fanpage (which is quite consistent with the world map that Google Insight draws for Nokia)

Note: I adjusted the contrast of the generated image, as the full color scale was not used well by the visualization.

And this is a list of related fanpages, i.e. a list of pages that Nokia Fans are more likely to be a fan as well:

Maybe even more telling might be to chart the Fan Pages which people are less likely to be subscribed to. We will see where Peter Warden is going to develop that tool.

“UNSD’s core mission is to advance the development of the global statistical system and promote the dissemination of statistical information. This database service is part of a project launched by UNSD in 2005, called “Statistics as a Public Good”, whose objectives are to provide free access to global statistics, to educate users about the importance of statistics for evidence-based policy and decision-making and to assist National Statistical Offices of Member Countries to strengthen their data dissemination capabilities.”

A couple of days ago I finally made my implementations of some common stochastic models for non-contractual consumer relationships available under the Apache 2.0 Open Source license: http://code.google.com/p/clv-master-thesis/

The code provides methods for estimating the according model parameters (via maximum likelihood), for calculating the likelihood of defection as well as forecasting future transactions on an individual level for the following models:

• NBD, Ehrenberg 1959
• Pareto/NBD, Schmittlein et al., 1987
• BG/NBD, Fader et al., 2005
• CBG/NBD, Hoppe & Wagner, 2007
• CBG/CNBD-k, Platzer, 2008

Additionally, I found time to upload my slide deck with a high-level, hopefully-easy-to-digest summary of my master thesis to slideshare.net: http://www.slideshare.net/mplatzer/my-entry-to-the-dmef-clv-contest

Anssi, my boss, talks openly about the disappointing quality of the N97. [I just love that finnish accent here at Nokia :-]

by the way: The Analytical Services Team, that I am part of, is further growing: Job Posting for Senior Data Analyst. Anyone interested in moving to Finland? :-)

On the occasion of the Winter Olympics Google Maps “Street View” went off-road and introduced “Ski Slope View” (at least for the ski resorts in Vancouver). Shooting these images must have been a bit more exciting than driving up and down every single road on this planet.

Microsoft, on the other hand, just recently revealed a new iteration of Bing Maps. And surprised many by not just catching up with Google, but actually surpassing it with many new never-seen-before features. It seems they managed to get back on their long-last path of innovation (at least regarding the Bing services). Watch this amazing TED talk (via the knallgang):

Not sure when it was the last time that Microsoft got standing ovations. Well, not sure when it happend for Nokia either :-)

Correlation does not imply Causation. On the other hand, “causation can only be inferred, never exactly known” [src]

Symbian^3

But somehow I have difficulties matching the images from streetview with the current scenes here in Helsinki. See for yourself:

As seen on Google Streetview:

Currently in Helsinki:

Notice a difference?

Here is another comparison:

versus

I already forgot that sometimes trees are green, and roads are grey :-)