[Scipy-organizers] Late Talk Submission

[Chris Moody]

Dear Conference committee,
Unfortunately, I missed yesterday's talk deadline -- my apologies, but
could you please still consider the following talk submission? I can no
longer submit this online and so I've copied the text below and included a
PDF of my talk if it helps!

Thank you so much for your consideration,
Chris Moody

Brief description:
Standard natural language processing (NLP) is a messy and difficult affair.
I'll speak about theory behind the word2vec algorithm and how Stitch Fix
uses this and LDA to digest and action on text. At the end I'll segue into
wildly creative extensions that carry word2vec from text and into the
domain of graph theory, translations, and sequence learning.

Detailed abstract:
Standard natural language processing (NLP) is a messy and difficult affair.
It requires teaching a computer about English-specific word ambiguities as
well as the hierarchical sparse nature of words in sentences. At Stitch
Fix, word vectors help computers learn from the raw text in customer notes.
Our systems need to identify a medical professional when she writes that
she "used to wear scrubs to work" and distill "taking a trip" into a Fix
for vacation clothing. Applied appropriately, word vectors are dramatically
more meaningful and more flexible than current techniques and let computers
peer into text in a fundamentally new way. I'll speak about word2vec and
related techniques and will try to convince you that word vectors give us a
simple and flexible platform for understanding text.

See this interactive blog post for more material:
http://technology.stitchfix.com/blog/2015/03/11/word-is-worth-a-thousand-vectors/

You can also see a preview of my talk here:
https://dl.dropboxusercontent.com/u/206140/temp/Word%20Vectors.pdf