How random is this sampling? The second highest country on the list is India (kind of understandable). The third highest is Sri Lanka. I'm not sure how it got that high. Also, UK rates below Paraguay and Nepal.
It's random in the sense that we let our crawlers just start at an arbitrary profile and stopped until we have about a million. Pretty random, thoughts?
Random graph walking can be random, but it is not always. The following paper on random walks on graphs as a tool for derandomizing probability algorithms gives insight on this, as well as the probability of a random walk on a RANDOM graph being a random sampling: http://www.cs.huji.ac.il/~nati/PAPERS/expander_survey.pdf
A lot of the location data is free text (ie. unstructured). Not easy to parse. Lat/Long more reliable, but for now, we're using AVG Lat/Long to determine country.Country data is a little skewed (for now).
Obtained like that it's not that interesting... A more interesting number would have been 'looking for love' but weighted on the size of the gender group.
That results in: 2.3% of Men looking for love and 0.37% of Women. These are very little numbers.
