"installing text extension with WordNet ( JWNL)"

cgkolar
cgkolar New Altair Community Member
edited November 5 in Community Q&A
Hello.  I have the text extension up and running with version 4.2.  I also have WordNet 2.1 installed on my XP box.  The docs mention a JWNL sample configuration file in the "samples" directory, but I cannot for the life of me find it.  I would appreciate any suggestions on taking this last step to link word net with RM.  Cheers,  --chris

Answers

  • Jens
    Jens New Altair Community Member
    Hi Chris,

    you can find such a configuration file in the 'config' subdirectory of the JWNL files you've downloaded.
    A basic description how to change this config file as available here http://jwordnet.sourceforge.net/handbook.html

    Unfortunately I didn't manage to use the WordNet operators like WordNetSynonymStemmer.
    I receive the error message
    Exception in thread "main" com.rapidminer.operator.OperatorCreationException: No operator description object given for 'WordNetSynonymStemmer'
    at com.rapidminer.tools.OperatorService.createOperator(OperatorService.java:531)
    Can anyone help me with that? I had that problem before when I installed the text-plugin, but I solved it by setting the rapidminer.home system property. Didn't help this time though  :-[

    Cheers
    Jens
  • land
    land New Altair Community Member
    Hi Jens,
    were did you take this operator from? It is not included in the Text Mining plugin and not in the core...

    Greetings,
      Sebastian
  • wotsiznamiz
    wotsiznamiz New Altair Community Member
    did you ever figure out how to get this to work?  I want to as well, but I can't figure it out...  I just started another thread (maybe i should've continued this one)...

    http://rapid-i.com/rapidforum/index.php/topic,777.0.html
  • Legacy User
    Legacy User New Altair Community Member
    Hi,

    the WordNet bridge is no longer supported by the TextInput plugin. Sorry for the outdated documentation...

    The main reason are the poor results it produced as JWNL did not support any part-of-speech or frequency information. So often ambigious words were not resolved correctly adding a lot of noise.

    However, there is an upcoming 2.0 release of JWNL with an extended API that looks very promising. So maybe it would be a good idea to use this new version to write WordNet operators for the current version of the TextInput plugin.

    As a point of departure, the code in the WVTool project could be used.

    Best regards,
    Michael
  • teadrinker
    teadrinker New Altair Community Member
    Doesn`t exist any solution to make the wordnet support work for the synonym stemmer step? Or isn`t wordnet supported at all in rapid miner 4.6 text plugin?
  • TobiasMalbrecht
    TobiasMalbrecht New Altair Community Member
    Hi,

    actually no, wordnet is not supported by the current version (4.6 as well as 5.0) of the text processing extension. Maybe we will re-add this feature somewhen in the future although this is not planned for the short-term.

    Best regards,
    Tobias