Text analysis on a collection of documents


Text analysis on a collection of documents

Hello everybody,


I have a huge csv file containing several columns, and, among them, there's one containing different texts, one per row. I need to compute text analysis (that means stemming, etc.) on each of these text documents separately, in order to (for instance) calculate relative frequence of some words, or more complicated estimation methods.


I tried out to select only the column of interest, and I went on but I do not know how to stem a collection of documents or an example set, and to use commands that are able to systematically do the same operation on these different texts, without merging them.


Could you please give me a hint on how to proceed? I hope that was clear.


Thank you very much,


Elite II

Re: Text analysis on a collection of documents

seems like you should use the "Data to Documents" operator in the Text Processing extension.  You're likely going to want to only do one attribute; check the "select attributes and weights" checkbox and in Edit List, select your text attribute and leave the weight to 1.0 (doesn't really matter).  That attribute will now be a "document" and you can use any of the Text Processing operators (e.g. stemming).




Scott Genzer
Certified RapidMiner Analyst
Genzer Consulting