In many occasions there is a need to identify keywords representing a text or a corpus of texts, so to classify or summarize it.
We achieve this result by using a techique called Non-negative Matrix Factorization (NMF) on the matrix Texts x Words, that allows us to identify topics and find out the keywords the represent each topic in a not ambiguous way.

You can test the results playing with the two texts below:

Like it? Share it! --- Facebooktwittergoogle_plusredditpinterestlinkedintumblrmail