[R] text mining - text comparing

[Matevž Pavlič]

Hi all, 

 

I'll try to explain what i would like to achieve. 

I have two problmes that i would need help on if someone has a clue. 

 

 

1.)    I have a TXT file containing two fields : USCS and Description.

 

For each field of USCS I have a field Descrition that contained a lot of words that describe that particular USCS type. What i would like to do is tomine the text  using tm package in order to find which words in Description filed are the most frequent for each USCS field. 

 

Now i don't think i will have problems with that part, but the problem is importing the data.  The thing is that there is areound 300 different USCS - Descritption  combinations which is of course to much to sort out by hand. I would have to create a Corpus of around 300 texts which I could later anylize.  Here is where i get stuck. I can not find a way to import the data in a Corpus so that i would have a text named after USCS value and containing strings (words) of Desription field. 

 

Attached (temp.txt) is a small dataset.

 

2.)    Second thing is about comparing text. I have some problems with typos in a text, so what i would like is to find a words that are similar (but spelled incorrectly). Similar that when typing in google engine, you get prposed words. Has anyone had any experiance in that? 

 

I hope i explaine ok, otherwise i'll try again, 

 

Tnx, m

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: test.txt
URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20110525/46b0af08/attachment.txt>