Hi R users,
I am using R for text mining and natural language processing, and would like
to accomplish a task that requires the determination of the distance between
two words.
The idea is to find how close are words with reference words like
"bad",
"good" and such, for use as a naive sentiment analyzer.
I am aware of Wordnet implementations in Python and Perl, and have also
grappled with calling Perl modules within R but my search for a pure R
implementation continues. The Wordnet package in R is great for other
lexical tasks (finding synsets, etc.) and currently does not support the
distance/similarity computations.
Would it be worth the effort to re-create the structure of Wordnet in R and
then query it? Or am I better off learning more about rJava and then calling
mature Java implementations within R for my purposes?
I understand that a trivial solution may not be near, but I would appreciate
any ideas, thoughts and suggestions that may take me closer to a feasible
solution.
Thank you very much for your time and patience.
Regards,
Harsh Singhal
[[alternative HTML version deleted]]
