This is a (very rough) prototype of a text editor that displays lexical information of the text entered. Each word in the text is colored based on its part-of-speech (blue = nouns, maroon = verbs, orange = determiners, etc.). Additionally, hovering over a word will highlight other similar words in the text. The more purple the highlight, the more similar the two words are. If the highlight is blue, then the words are synonyms.
The python package nltk powers the POS tagging and word similarity. The text area is a (customized version of) a HTML Canvas text editor called carota.