Text::MeCabとTermExtract::MeCabでキーワード抽出

MeCab + TermExtract.pl

sub morphological_analysis_by_mecab {
    my $decoded_str = shift;
    my @array_terms;
    my $mecab = Text::MeCab->new({
	userdic => 'hoge.dic',
				 }
	);
    my $str = encode('euc-jp', $decoded_str);

    my $node = $mecab->parse($str);
    my $text = '';
    for( ; $node; $node = $node->next ) {
	if ( defined $node->surface && is_ideal_word($node->feature) ) {
	    $text .= $node->surface . "\t";
	    $text .= $node->feature;
	    $text .= "\n";
	    push @array_terms, decode('euc-jp', $node->surface); 
	}
    }
    $text .= "EOS\n";
    return { mecab_terms => \@array_terms, mecab_output_string => $text };
}
sub get_terms_by_term_extract {
    my @array_terms;
    my $mecab_output_string = shift;
    my $term_extract        = new TermExtract::MeCab;
    my @noun_list           = $term_extract->get_imp_word($mecab_output_string, 'var'); # 入力が変数

    foreach (@noun_list) {
	# 日付・時刻は表示しない
	next if $_->[0] =~ /^(昭和)*(平成)*(\d+年)*(\d+月)*(\d+日)*(午前)*(午後)*(\d+時)*(\d+分)*(\d+秒)*$/;
	# 数値のみは表示しない
	next if $_->[0] =~ /^\d+$/;
	
	# 結果表示
	push @array_terms, decode('euc-jp', $_->[0]);
    }
    return \@array_terms;
}