User:Fako Berkers/project2

Sniff, Scrape, Crawl

Wicked Wiki

I started the year with a scraper/tool that can investigate persons on Wikipedia. At the moment it can find collegues based on simularities in categorization. This is a great tool, because if you find one person interesting you are likely to appreciate the others. Investigating Ghandi for instance delivers some unknown Indian rulers and peace activists. I'm still working on improvements. You can read more here.

Anecdote: from WW to AA

When I was on my way to Berlin to visit Transmediale I was thinking about my Wicked Wiki idea. The program was slow, but functional and I was looking for a next step. For me this was in visualizing results. Something that concerned me was that a lot of people on Wikipedia disappear into the unknown. That's why a wanted to make superstars out of people that were on Wikipedia but not famous.

When I got back from Berlin the results of my test subject (Slavoj Zizek) were poluted with "German Nachste Superstar", the German Idol program on television. Quite baffeled by this coincedent I started to work on a program that visualized the battle between Zizek and Lombardi (one of the German Idol participants). This grew into the much bigger tool that is now Attention Arena

Attention Arena

I'm interested in how internet regulates our attention. Some subjects may be very interesting, but because the majority doesn't watch these subjects they fade away on the internet being replaced by what others are watching. Attention for some interesting subjects will disappear because of this.

I started to visualize this by putting two YouTube videos on top of each other fading the least popular video. This worked well with a video from Zizek and Lombardi, a philosopher and teenage superstar respectively.

Spurred on by this nice result I started to work on a tool that does the following:

• It allows a user to create a list of Youtube videos in Django Administration.
• These videos get downloaded automatically and converted for manipulation purposes if you run a script.
• The videos can be shown in OpenFrameworks.
• With Python you can save preferences for viewing in Django like: size, position, transparency and number of videos

My plan is to create a pyramid of six videos. The top video will be the most popular one and the lower ones less popular. The lower the video is on the pyramid, the more faded the video appears. This setup shows what is popular and what is not. It reveils where the attention of the public is heading and what slowly will be forgotten.

To write an algorythm that creates this pyramid will not be too hard. It is a challange to find a good collection of Youtube videos that strengthen my concept behind the form. This is where the work is at now.