Snowball Stemmer Usage

Question

I'd like to use the stemmer here for merging word counts.
http://snowball.tartarus.org/download.html
The page has a download link, but I'm not sure how to integrate the files into my eclipse project
Its not just a jar to drop into my lib folder, its a file system. Does anyone know of some documentation explaining this, as I didn't see any on the website.
(As in, what do i import, how do I call it etc..)

Rurik

The snowball manual and The snowball how to run it – Creese

i've read both of those, but the second one says how to run it standalone with java, not how to import into a project and the first one doesn't touch on real setup – Rurik

Answer 1

Build the jar file and add it to your Build Path.

Details:

• Download the tgz with the code from here http://snowball.tartarus.org/download.php
• Uncompress.
• Go to libstemmer_java directory and read README.
• Follow instructions to compile (using javac).
• You might have to correct or remove java/org/tartarus/snowball/ext/frenchStemmer.java because it has an error and doesn't compile.
• Create jar file: Go to libstemmer_java/java directory then jar cvf libstemmer.jar *
• Add libstemmer.jar to your Build Path (in Eclipse: Project-Properties-Java Build Path-Libreries Tab).

Then you can use the stemmers doing something like:

import org.tartarus.snowball.ext.spanishStemmer;
...
spanishStemmer stemmer = new spanishStemmer();
stemmer.setCurrent("torero");
if (stemmer.stem()){
    System.out.println(stemmer.getCurrent());
}

Multicellular

I ran into a compile error so I took out all but the English language packs and the compiling went perfect. Thank you for actually answering this question and not telling someone to RTFM. :) – Astrakhan

i am having an unusual problem . i got my string in a variable called "word" like word="torero"; and when i pass this variable to to the stemmer it wont work. eg stemmer.setCurrent(word); stemmer.stem(); System.out.println(stemmer.getCurrent()); .it wont get stemmed. tell me what i am doing wrong here . – Donelladonelle

No @JunaidShirwani. It sometimes return wrong output. Same thing happening to me. – Knockabout