view release on metacpan or  search on metacpan

README.pod  view on Meta::CPAN

=encoding utf-8

=head1 NAME

AI::Classifier::Japanese - the combination wrapper of Algorithm::NaiveBayes and
Text::MeCab.

=head1 SYNOPSIS

    use AI::Classifier::Japanese;

    # Create new instance
    my $classifier = AI::Classifier::Japanese->new();

    # Add training text
    $classifier->add_training_text("たのしい．楽しい！", 'positive');
    $classifier->add_training_text("つらい．辛い！", 'negative');

    # Train
    $classifier->train;

    # Test
    my $result_ref = $classifier->predict("たのしい");
    print $result_ref->{'positive'}; # => Confidence value

=head1 DESCRIPTION

AI::Classifier::Japanese is a Japanese-text category classifier module using Naive Bayes and MeCab.
This module is based on Algorithm::NaiveBayes.
Only noun, verb and adjective are currently supported.

=head1 METHODS

=over

=item C<< my $classifier = AI::Classifier::Japanese->new(); >>

Create new instance of AI::Classifier::Japanese.

=item C<< $classifier->add_training_text($text, $category); >>

Add training text.

=item C<< $classifier->train; >>

Train.

=item C<< my $result_ref = $classifier->predict($text); >>

Test and returns a predicted result hash reference which has a confidence value for each category.

=item C<< $classifier->save_state($params_path); >>

Save parameters.

=item C<< $classifier->restore_state($params_path); >>

Restore parameters from a file.

=item C<< my @labels = $classifier->labels; >>

Get category labels as an array reference.

=back

=head1 LICENSE

Copyright (C) Shinichi Goto.

This library is free software; you can redistribute it and/or modify
it under the same terms as Perl itself.

=head1 AUTHOR

Shinichi Goto E<lt>shingtgt @ GMAIL COME<gt>

=cut