Slide 1

Slide 1 text

Visualizing word coincidences Piotr Migdał, PhD data scientist @ deepsense.io
 
 http://p.migdal.pl/ Story, Character & Dialogue Visualization with D3 Bay Area d3 User Group 14 Mar 2017, Los Gatos (CA)

Slide 2

Slide 2 text

viz: https://mikolajczyz.com/general/research-addictions-results/#cloud repo: https://github.com/stared/nalogi-viz

Slide 3

Slide 3 text

viz: http://p.migdal.pl/tagoverflow/?site=english&size=32 repo: https://github.com/stared/tagoverflow/

Slide 4

Slide 4 text

viz: https://lamyiowce.github.io/word2viz/ blog post: http://p.migdal.pl/2017/01/06/king-man-woman-queen-why.html word2vec analogies

Slide 5

Slide 5 text

viz: https://lamyiowce.github.io/word2viz/ blog post: http://p.migdal.pl/2017/01/06/king-man-woman-queen-why.html stereotypes

Slide 6

Slide 6 text

https://twitter.com/cesifoti/status/818672972743450624

Slide 7

Slide 7 text

https://news.ycombinator.com/item?id=13346104 word2vec + AD&D classes

Slide 8

Slide 8 text

Thank you! Piotr Migdał, PhD [email protected] http://p.migdal.pl click: https://lamyiowce.github.io/word2viz/ (yes, it’s interactive!)