What does StackOverflow’s personalized prediction data think of you?
While reading David’s “One year as a Data Scientist at Stack Overflow”, I’ve learned about one of their products: Providence (a system for matching users to jobs they’ll be interested in).
Not sure why I’ve never heard about it before, but anyway, I was curious to see what it learned about me.
Luckily that wasn’t very hard, since they offer a way to download your personal providence data.
You get a nice JSON file, with TagViews
and InterestingTags
being the interesting part with the technology/scores.
We could create a small app to parse and sort, but I’m lazy so here’s a quick bash one-liner for a top 10 for TagViews
:
grep '^ \"' "Stack Exchange personalized prediction data 2016-06-21.json" | sort -nr -t ":" -k2 | head -n 10
And for InterestingTags
:
grep '^ \"' "Stack Exchange personalized prediction data 2016-06-21.json" | sort -r -t ":" -k2 | head -n 10
For me, the results are pretty spot on.
"java": 214,
"android": 88,
"gwt": 87,
"javascript": 43,
"json": 26,
"mongodb": 25,
"linux": 24,
"python": 22,
"bash": 18,
"jquery": 17,
and
"nginx": 0.529183447360992,
"android": 0.402329504489899,
"java": 0.278640985488892,
"linux": 0.0971642881631851,
"javascript": 0.0780074372887611,
"networking": 0.0450629182159901,
"jquery": 0.0448367521166801,
"ubuntu": 0.0409384779632092,
"html": 0.0403672903776169,
"php": 0.0382767021656036,
Nice!
Read other posts