Wordmesh: Using t-SNE and word2vec to generate clustered wordclouds

Sometimes I find wordcloud visualization to be very trivial. It has always bugged me to see many people using wordcloud just because it can look cool sometimes. Using font colors, size and clustering is a great way to visualize meaningful dimensions.

Since most of my analysis is document level, I create content and relation based graphs of documents and use Gephi to visualize word and document clusters in a meaningful and beautiful way.