KinoSearch1::Analysis::Stemmer - reduce related words to a shared root
my $stemmer = KinoSearch1::Analysis::Stemmer->new( language => 'es' ); my $polyanalyzer = KinoSearch1::Analysis::PolyAnalyzer->new( analyzers => [ $lc_normalizer, $tokenizer, $stemmer ], );
Stemming reduces words to a root form. For instance, "horse", "horses", and "horsing" all become "hors" -- so that a search for 'horse' will also match documents containing 'horses' and 'horsing'.
This class is a wrapper around Lingua::Stem::Snowball, so it supports the same languages.
Create a new stemmer. Takes a single named parameter,
language, which must be an ISO two-letter code that Lingua::Stem::Snowball understands.
Copyright 2005-2010 Marvin Humphrey
See KinoSearch1 version 1.01.