Bookworm on vector space models
« previous post | next post »
A couple of great posts by Ben Schmidt at Bookworm: "Vector space models for the digital humanities", 10/25/2015; and "Rejecting the gender binary: a vector-space operation", 10/30/2015.
Update — A quick experiment by a Penn grad student, which confirms that somewhat plausible things emerge from fairly small and fairly noisy datasets…
Jerry Friedman said,
November 1, 2015 @ 5:27 pm
Very interesting!
It looks like the kind of thing that could be overstated by journalists in the way that you often criticize. We'll see.
[(myl) The basic results (word2vec, GlovE, Eigenwords) have been around for a couple of years. And Schmidt's description seems to me to strike the right balance between enthusiasm and skepticism.]
Rubrick said,
November 2, 2015 @ 4:32 am
Very interesting indeed, and quite accessible to this non-linguist, non-R-programmer.
Daniel Barkalow said,
November 2, 2015 @ 7:41 pm
I laughed at #25 on the list of gendered synonyms in "Rejecting the gender binary".
MikeA said,
November 2, 2015 @ 7:44 pm
Did I miss a calendar reform? That second post was apparently made
on Octodecember 30th, 2015.
[(myl) You mean "F4ember", right? Anyhow, it's fixed now.]