Convenience functions for simple workflows

Having started to use this package coming from `quanteda` in R, I feel some convenience features could greatly streamline common workflows:

* Top features in corpus/subset of corups/document
* Filtering DTMs by eg. proportion of documents that contain a word, or a regex pattern
* Clearer documentation around `DocumentTermMatrix` vs `dtm()` functions, eg. on the ordering of terms
* Coding a document term matrix with e.g. a sentiment dictionary (cf. ` quanteda.dictionaries`) package inR

I'd be happy to contribute PRs for these, if any/all are desired functionality. That said, I'm quite new to the ecosystem so if I'm missing something do let me know!