Check out the blog post at eptikar.com/blog
We used the awesome data availabe at data.stackexchange.com, using SQL to fetch related subsets of data.
Users's LOCATION is an open feild in Stackoverflow database and there are about 47,835 unique location by the time of makaing this study.
Using nested LIKE
conditions we get possible matches of the location, for example to pull users from Egypt we would have to do:
# Fetch users from Egypt
SELECT * FROM users WHERE Location LIKE '%Egypt%'
OR LOWER(Location) LIKE '%egypt%'
OR LOWER(Location) LIKE '%cairo'
OR LOWER(Location) LIKE '%alexandria%'
OR LOWER(Location) LIKE '%pORt said%'
OR LOWER(Location) LIKE '%giza%'
OR LOWER(Location) LIKE '%fayyum%'
OR LOWER(Location) LIKE '%suez%'
OR LOWER(Location) LIKE '%al mansurah%'
OR LOWER(Location) LIKE '%mansoura%';
OR LOWER(Location) LIKE '%sharm el sheikh%'
quiries.sql contains the sql used to match users of the Arab countries
P.S:
Of course this could be refined to give more accurate results
We used R to make sense of the data using simple visulations, results are in the graphs directory
###CREDIT:
- Rayan
- Nimir
###LICENSE: GPL V 2.0