Mike Schilli
and 1 contributors


Text::TermExtract - Extract terms from text


    use Text::TermExtract;

    my $text = { Hey, hey, how's it going? Wanna go to Wendy's 
                 tonight? Wendy's has great sandwiches." };

    my $ext = Text::TermExtract->new();

    for my $word ( $ext->terms_extract( $text, { max => 3 }) ) {
        print "$word\n";

    # "sandwiches"
    # "tonight"
    # "wendy"


Text::TermExtract takes a simple approach at extracting the most interesting terms from documents of arbitrary length.

There's more scientific methods to term extraction, like Yahoo's online term extraction API (but you can't have it locally) and the Lingua::YaTeA module on CPAN (which is so poorly documented that I couldn't figure out how to use it).

So I wrote Text::TermExtract, which first tries to guess the language a text is written in, kicks out the language- specific stopwords, weighs the rest with a hand-crafted formula and returns a list of (hopefully) interesting words.

This is a very crude approach to term extraction, if you have a better method and want to include it in Text::TermExtract, drop me an email, I'm interested.




terms_extract( $text, $opts )

Goes through the text stringin $text, extracts the keywords and returns them as a list.

To limit the number of words returned, use the max option:

    $extr->terms_extract( $text, { max => 10 } );
exclude( $array_ref )

Add a list of words to exclude. The words listed in the array passed in as a reference will never be used as keywords.

    $extr->exclude( ['moe', 'joe'] );


Copyright 2008 by Mike Schilli, all rights reserved. This program is free software, you can redistribute it and/or modify it under the same terms as Perl itself.


2008, Mike Schilli <cpan@perlmeister.com>