Datascientific Search Image

I’m currently experimenting with HTML5 Canvases for a project. Among other things, I created the following “animation”. Here colourful dots are painted onto a canvas at random positions, only one word is left out. The word is randomly drawn from all words on the current page. To find out which word is hidden at which position in the canvas, you have to watch the animation for quite a while..

The animation will only start with activated JavaScript.

While you watch the animation you are actually doing data science! You observe data points and try to abstract from them to find a pattern :)

Admittedly, data science is still a bit differently. But in both cases, individual data points transport relatively little information. Sometimes the information is misleading or distracting: In the animation above, for example, the colour of the points has no meaning at all, it’s just to look pretty… However, when many data points have come together, a pattern emerges and you can first guess where the word is located in the picture, and with further data you can recognize parts or later the entire word. It also becomes clear that the surface does not have to be completely painted - the word can be clearly recognized much earlier.

Data is the oil of the 21st century they all say… Even if I find the comparison a little … not so suitable, the idea behind it is of course not entirely wrong. Data plays an increasingly important role in our lives, just ask Google! When you watch a video on YouTube, Google only knows a little bit about you. If you do this more often, Google will know a little more. If you use Google for every online search, manage your emails via Gmail, and surf the Internet without a tracker-blocker, Google will soon be able to assess you better than you could ever do yourself… The same applies, of course, to all the other small and large players like Facebook or Amazon, who want to track us at every turn. That’s why I plead for data economy: Don’t be so careless about your privacy!

For “copyright” reasons there is no download button for the automatically generated image. So you can’t take it with you, but have to come back if you want to see it again - each time with different data points, of course! So it’s worth coming back every now and then ;-)

The code of the animation is moderately commented and can be downloaded from this webserver.