Uploaded on 2015-04-16 by HYoussefA

The photo shows two buildings on the campus of UPenn in Philadelphia during a snow storm. I chose it because it illustrates an issue I have with big data. 1) Visible information: buildings, clearly not of the 20th century. Cars parked on the side of the street, bike racks. The street is clear of traffic (these cars are all parked), so the storm is strong enough to stop traffic. 2) Invisible information: because we are limited the information that can be extracted from one photo, this is necessarily limited. First, the function of the buildings might be deduced: there is actually a campus map in the photo, towards the bottom left (if you squint hard enough...). Secondly, the fact that there's a snow storm, and the lights are all out, suggest the occupancy of the buildings to be very small, maybe even no one is in. 3) Knowledge for planning: severely reduced building occupancy during extreme weather means that building managers can plan to reduce power supply to these buildings when the forecast predicts extreme weather. The exercise is a bit contrived because we have to extract or infer information from one picture. But the point from my choosing this picture is that livable cities probably have an element of surprise to them that is hard to quantify. To me, this is a charming part of town, and a calming scene. Part of its charm is its mystery, and the fact that it resists analysis to some point... If those cars weren't in the frame, one would be justified in thinking this to be a 19th century scene..(assuming colored pictures were around then :) ) [1]: https://edxuploads.s3.amazonaws.com/1429158069965952.jpg