25. februar 2010

My familiar strangers this morning

I have been experimenting with data logging on Pachube and the concept of the "familiar stranger" (people you coincidentally create social networks with due to same habits or routines). I believe there is a big potential for interesting analytical use of the data.
This morning I logged my surroundings going from home to work. The thumbnails show two graphs, one for new strangers (top)  and one for familiar strangers (bottom). In this case "familiar" is concidered to be people that have been around for more than one minute.
Overlaying the two graphs I'll try to explain what caused the fluctuations in the data.
Vertical lines are five minutes apart, blue is for strangers and red is for familiar strangers.
  1. Leaving kindergarden and walking towards subway stop. I was alone.
  2. Approaching the station, buying a coffee at the subway station and a seven minute wait at station. People come and go and on the platform, but there are a few people waiting with me.
  3. Entering rail car, lots of people around. I guess the second blue spike is the next station. 
  4. The rail car fills up and the same people are with me until I approach my transfer station (15 min journey). See how the two blue spikes adds up to the height of the red graph, suggesting that the people from the two first spikes are within 5-10 meters.
  5. Changing line, meet new people, but the rail car is less crowded than on the other line. 
  6. Leaving the subway and passing through an large office builidng.
  7. At work, quite a few familiar strangers that i suspect will stay for the duration of the day.
I find it interesting to see how the blue spikes on some occations are followed by a longer red surge and sometimes not. This tells me, just from the data that some occations you move past people and sometimes you stay with them. Which is what you would expect!

1 kommentar:

  1. I just realized the concept of familiar stranger. This would be very brainstorming article. Really great to read your thesis.

    SvarSlett