Would it be possible to create a tool for using RAKE? 1. input a list of PDFs, of a certain type, or of any type 2. extract text and apply RAKE 3. output keywords for each document, or each type of document That way it would be useful to read academic papers