It's interesting to show the VP and 45, but as their speeches and public messages are probably mostly written BY the senior staff the correlation is highly suspect.<p>Also, just because you can do something doesn't mean you should.
No Kellyanne Conway as an option? That was my bet.<p>Pence uses the word "lodestar," Mattis loves to say "first principles." The ending (reach across aisle) reads a lot like McCain speak. ...Of course the 42D Chess move is to make it look like anyone near Trump could have written it.<p>It's also possible it was penned by two or three people, especially if it was a group that formed a pact to save their jobs/face as un-elected bureaucrats, as a hedge if the midterms go poorly and they still want D.C. jobs, etc.
Flagging this because doxxing is not something this site should promote, no matter how high-profile the post, nor how politicized the content. No matter how interesting the techniques used, this thread is potentially as dangerous as Reddit's witch hunt for the Boston bomber. [0]<p>[0] <a href="https://www.reddit.com/r/MuseumOfReddit/comments/1iv343/the_boston_bombing_debacle/" rel="nofollow">https://www.reddit.com/r/MuseumOfReddit/comments/1iv343/the_...</a>
Does this filter out retweets, and/or the text of retweets? It would be useful to see the raw data it's built from—the code specifies a 'data/mat.rds' file, but there's no data directory in the repository, and I don't have R installed / any familiarity with it to really follow details about the code.
The NYT knows better. They know that you can unmask someone by finding correlations in style. They most likely edited the content heavily to avoid this.<p>Besides, correlations with tweets probably wouldn’t mean much as tweets are not written the same way as regular text and not by the same person anyway.
Points to a senior speech-writer. Someone who has written speeches for both Pence and POTUS - but probably more speeches for Pence... might be an op-ed directed by Mattis but my money is on the speech-writer going rogue.
I'm surprised by this result, given the fact that the op-ed is compared to tweets and DJT turns out to be one of more highly correlated members of the group. Based on what I've seen posted on his Twitter account in the last two years and the way he's handled himself in public since taking office, it's really hard to see how this op-ed came from him personally.
Very relevant: <a href="https://twitter.com/TheRickWilson/status/1037512349488492546?s=19" rel="nofollow">https://twitter.com/TheRickWilson/status/1037512349488492546...</a>
A lot of speculation on Twitter is focused on the word lodestar which has previously been used by Pence.<p><a href="https://twitter.com/jamescdownie/status/1037441482469335043" rel="nofollow">https://twitter.com/jamescdownie/status/1037441482469335043</a>
Dave Robinson has a nice blog post using a similar analysis:<p><a href="http://varianceexplained.org/r/op-ed-text-analysis/" rel="nofollow">http://varianceexplained.org/r/op-ed-text-analysis/</a>
The high score of VP and POTUS is surely reflective of their speech writers... isn't this backed up by the fact that
VP has .8 and mike_pence has -.05?<p>Regardless, if throw out POTUS and VP, we are left with Pompeo and Ross as the two highest.