Use the mouse to click on a dot and see the word plotted. Click a label to hide it again.
Color indicates polarity based on simple word labeling from the AFINN wordlist. It may be that context in this dataset affects polarity :)
forked from arnicas's block: TSNE Plot of Yelp Reviews
https://cdnjs.cloudflare.com/ajax/libs/d3/3.5.6/d3.min.js