"Dictionary based analysis"

mlctu

Hi!

Is there a way to use RapidMiner to perform dictionary-based analysis of document collections?

In particular I'm interested in term frequency and other statistics to be applied to term occurrences in the documents, where the terms of interest are provided by the user, already classified in one or more user-defined category lists (dictionaries).

Thanks for your help!
Giulio

land

Hi,
yes this is easily possible with the Text Processing Extension. You can simply use the Dictionary based filtering to remove all uninteresting words.

Another way around would be to first count all words and then postprocess this word list using the "WordList to Data" operator.

Greetings,
  Sebastian