Last night I read up on Latent Dirichlet Allocation for text mining on unstructured text, and now I'm seeing applications everywhere. Time to go try it on a real dataset...
It seems I'll be out of job in January, so I ask you for suggestions.
I'm a PhD in #psychology and licensed #psychologist. I do work psychology, psychological assessment and narrative #psychotherapy. I mainly teach critical social psychology, personality psychology, organisational psychology, research methods, statistics, psychometrics, discourse psychology, #phenomenology, and history of psychology. And I use :rstats:.
So I realize I've been thinking of social networks as graph-based and set-based.
Graph-based means you follow other users and get a feed of their posts. (E.g. Facebook, Goople, birdsite, here).
Set-based means you are either in or out and if you're in, your feed is the same as everyone else who's in. (E.g. subreddits, mailing lists, USENET groups, web fora.)
"I'm waiting for the DataFrame to populate" is the new "I'm waiting for tests to finish" is the new "I'm waiting for it to compile" is the new "have you finished booting it yet?"
@fitheach@alcinnz can't remember the name, but somebody made an absolutely great point that selecting the people into institutions is not the only important function of elections.
Another function, *as important*, is building trust in the process of peaceful transition of power.
It's hard to trust something one does not understand.
Paper ballots are simple, verifiable, understandable. Electronic voting is anything but.
Loving Julia Silge & David Robinson's book on Tidy Text Mining - as a true acolyte of the One True (Tidyverse) Way, I approve of this approach compared to what I've seen of "tm" and similar (although I'll likely have to use "tm" for prediction stuff, it seems).
Also it makes text work easy to pipe into ggplot2, glorious...