The Belgian elections are closing in fast. Political pollsters and classic media worked hard to get some first impressions of the Belgian political mind. In addition, creative agencies launched campaigns that also observe a thing or two online. Fact is that Obama did have his online business pretty well covered, and it might have had a significant impact on his results. This might trick you into thinking buzz could represent voting interest, but that wouldn’t be a solid thought process.

While buzz is an interesting aspect to look at, it often equals to push and is generated by a target group that is most likely not that representative.

 

 What I would like to see, is data showing what political information Belgian internet users actively pull in just before the elections. In this case, pull would translate to search queries.

 

 

 

No idea. The provided tools are geared towards advertising – not politics. But a similar experiment was done on predicting the next American idol. The analysis proved irrelevant as the other guy seems to have won.

 

 

A big bias when merely analyzing political party search volume is the real voting behavior. As Pietel already pointed out: people might cast votes from sympathy towards a political personality. Party search volume would then become obsolete. A new analysis with all politicians and their working parties would be required to make the results more valuable again. Since I don’t have time for that, I won’t go into this.

A second bias would be generated buzz. When search volume is generated because of news, fraud, or buzz in general around a political party, this can’t be seen as a representation of voting interest.  

 

 

 

Let’s anticipate on that. Above graph shows proof of biased search volume: the search spike clearly yields the news spike on both political parties. When going back in time to what happened, we recall a defecting politician between these two parties. The search volume was biassed at that point and probably doesn’t reflect voting interests.

 

Last but not least: the users that do make an effort on doing some online research are probably not representative for the Belgian collective mind.

 

 

 

 Just because we can. Plus, voting is mandatory in Belgium - and with an internet penetration of nearly 70% there somehow still might be a significant amount of people researching their parties of interest shortly before the elections. Next to this, many people still seem undecided.

Second, I’m assuming users run these searches from a frame of interest - not entertainment. If we look at the data from the American Idol show, we see big search spikes after a contestant gave a good performance. So that could mean people might have been coming from a frame of entertainment at those points, and not genuine voting interest.

 

So based on all these assumptions, don’t take anything that follows serious.

 

 

 

Limited to Flemish observations, these were the main things that came out:

• CD&V stays the biggest party.
• LDD won’t grow or even loses
• Green party breaks in
• SLP doesn’t stand a chance

++ Update: The VRT just published new poll results. Seems like they don't agree with the second statement.

 

 

Some graphs with search volume:

 

 

 

 Note that this in no way can be compared to eventual party sizes.  In fact, these graphs are irrelevant as it more interesting to look at breakout and rising search terms to see if a party is gaining interest lately. For some reason, the small SLP is doing something weird here. Along with LDD in a tie it racks up the breakout searches. Both have 3 breakout search terms compared to the other parties. (Image below only shows breakouts in comparision with VLD)

 

 

 

Next to this there is an unexplainable search peak for this minor party at the beginning of May. I’ve pulled it through the trend analysis and Google news search, but can’t find out why that happened. Nevertheless; since such peaks are fishy and probably biassed, we better discard it.

 

 

What else is notable, is that the VLD ranked most top searches compared to other parties. The CD&V didn’t even appear in the rising searches, but kept it more or less steady to what it has always been doing. Don’t get this wrong: CD&V and NVA still account for most search volume, but only when combined. This year they will go as separate parties to the elections, and thus VLD is taking over as the party with most search volume and top searches compared to other parties. Noteworthy fact is that they had the same search behaviour during the previous elections and did lose. But this time, it seems like CD&V and NVA will have to share some interest and the VLD would benefit from that.
 

 

 

 So to sum up what appears to be going on with political related searches within Belgium:

• Green indeed does move forward. (2 breakouts)
• Not CD&V but VLD generates most volume
• LDD still gains in popularity (3 breakouts = most)

 

 

 

As said, seen all the biasses and the previous experiment: not at all. Another big problem is that Google doesn’t provide absolute search values. So that leaves us guessing towards the numbers they are normalizing. Because of this, we can’t even say anything about statistical correctness of the representation of the search volume. Right now the graphs seem very jumpy, what might indicate low absolute numbers.

I’ve ran sort like analysis on the elections of June 10, 2007 as a reference point, but didn’t get solid matches in terms of party sizes with the eventual result back then - although it was close.

Two things I maybe could have distilled from the data between June 3 and June 9 in 2007 were:

But let’s just cast our vote and see afterwards.

 Project Natal was introduced yesterday at the E3 conference. Impressive demo gear, but some might look at it critically and wonder about the input. More specifically about the seamless conversion of image to body movement. If you’re interested in my wild guess on the technology side, carry on – but don’t take it for granted.

 

 

 

When I saw it, the first thing that came to mind was the research of Tokyo University engineer Tsuyoshi Horo. He interfaced his robot by gestures.  What I found particularly interesting was the way he used and processed his input from the cameras. The output was based on Human Volume Reconstruction. This means the system perceived the user as a virtual object composed out of little virtual cubes.

 

 

Technically there was an array of cameras to detect body movement. This resulted in a real-time 3D volumetric model of himself that consisted out of these small virtual cubes. By analyzing these cubes as data, gestures could be extracted. From there on its all math and probability calculations’ mapped on the model of a human body. The video below shows the ropes:

 

 

Fact is that Human Volume Reconstruction requires a whole lot of processing power, but that might explain the extra processor in the Natal hardware.

Key about volume calculations is not that much image quality, but having objects that are physically separated. When they are separated, they can fairly easy be calculated and processed. This might be a limitation for using it as input from a game environment since people tend to sit closely together in a couch.

 

Another limitation for Natal with Horo’s setup would be the need of an array of cameras. A big no-no for the average Joe’s living room. My bet is that Microsoft seems to have that covered by strongly refining the technology, and using the RGB camera for just height (y) and width (x) perception. A depth sensor would cover the Z-dimension.

 

Sounds logical, but the big problem with all depth sensors (infrared, ultrasound, etc) is that they generally tend to work in very fixed line of sight, and thus making them pretty useless for Human Volume Reconstruction. So, that kind of puts this theory in thin air again.

 

Or maybe, just maybe, they are making use of improved hardware like the one developed by Elliptic Labs. This Norwegian company creates touchless 3D interfaces (based on depth) for a while now. But the only problem with their depth perception is that the user can’t provide input if he is more than 1,5 away from the perceiver:

 

 

So, however it would work, it will be an interesting revelation.

 

 

First thoughts: red eye effect. In combination with the same camera that measures the X and Y volume, the users’ eyes could be tracked within the image, even in dark environments. This would be done by constantly sending out infrared light. The users’ pupils would reflect this, (also known as the red light effect) and show up in Natal as bright dots. 

 

There is a company that uses this concept to measure ad exposure based on how many times people look at an ad. Above you can see an image of the hardware they put around the ad, but this is more expensive than it looks.

 

 

Last year I did a project for the university based on a speech interfaces. We prototyped some interfaces for cars (controlling the radio etc) and for a Pizza ordering telephone hotline. Since our software was open source, I’m quite sure that it wasn’t state of the art, but still..

The setup worked with flowcharts and actions were triggered by keywords. We user tested it with people with different English accents, and that worked quite well for simple things with limited answers. So keyword based speech recognition is something that would be possible. I do have serious doubts about natural language recognition though, but I don’t think that is needed for interacting with games. The user will probably have to be trained a bit in the use of keywords.

 

 

Most exiting will be to see how movement perception works – but I’m also most optimistic about accuracy of this one, seen recent research. If I had any concerns, it would be about the speech recognition and the overall cost of a system that combines all this technology. I can already tell: if it uses the above technology, it won’t come cheap. Natal input that eventually would cost more than the Xbox itself should not be a surprise.