I’m trying to set delimiters for WordTokenizer  as a parameter to StringToWordVector  which is in turn a parameter to FilteredClassifier. I have not been able to do this using a command line or an xml file for the options. This command line works without the –delimiters flag:

 

$ java -cp "c:\incoming\weka\developer-branch\weka.jar" weka.classifiers.meta.FilteredClassifier -F "weka.filters.unsupervised.attribute.StringToWordVector -C -L -T -I -S -W 2000 -N 1 -tokenizer \"weka.core.tokenizers.WordTokenizer\"" -W "weka.classifiers.bayes.NaiveBayesMultinomial" -t test.arff

 

How would you add “-delimiters <delimiter_string>” here to the value for –tokenizer, e.g. if the delimiter string is “ \\t\\r\\n”?

 

Thanks!

David Law