Charting the English Language...in pure Clojure
Alexander Mann at Clojure/conj 2016
In 2008, Collobert and Weston analyzed 2500 english words using deep neural nets. Since then, people like Joseph Turian have used technology such as t-SNE to visualize these results in a way that humans can actually comprehend (2 dimensions instead of thousands). This talk explores a naive Clojure programmer’s attempt to reproduce Turian’s word map of Collobert and Weston’s word feature vectors in pure Clojure from the raw sentences data set. Attendees will walk away with the contextual understanding and basic toolset to reproduce this study.