man Text::Ngram (3): Ngram analysis of text

SYNOPSIS

use Text::Ngram qw(ngram_counts add_to_counts);
my $text = "abcdefghijklmnop";
my $hash_r = ngram_counts($text, 3); # Window size = 3
# $hash_r => { abc => 1, bcd => 1, ... }
add_to_counts($more_text, 3, $hash_r);

DESCRIPTION

n-Gram analysis is a field in textual analysis which uses sliding window character sequences in order to aid topic analysis, language determination and so on. The n-gram spectrum of a document can be used to compare and filter documents in multiple languages, prepare word prediction networks, and perform spelling correction.

The neat thing about n-grams, though, is that they're really easy to determine. For n=3, for instance, we compute the n-gram counts like so:

    the cat sat on the mat
    ---                     $counts{"the"}++;
     ---                    $counts{"he "}++;
      ---                   $counts{"e c"}++;
       ...

This module provides an efficient XS-based implementation of n-gram spectrum analysis.

There are two functions which can be imported:

ngram_counts

This first function returns a hash reference with the n-gram histogram of the text for the given window size. The default window size is 5.

    $href = ngram_counts(\%config, $text, $window_size);

As of version 0.14, the %config may instead be passed in as named arguments:

    $href = ngram_counts($text, $window_size, %config);

The only necessary parameter is $text.

The possible value for %config are:

flankbreaks

If set to 1 (default), breaks are flanked by spaces; if set to 0, they're not. Breaks are punctuation and other non-alphabetic characters, which, unless you use "punctuation => 0" in your configuration, do not make it into the returned hash.

Here's an example, supposing you're using the default value for punctuation (1):

  my $text = "Hello, world";
  my $hash = ngram_counts($text, 5);

That produces the following ngrams:

  {
    'Hello' => 1,
    'ello ' => 1,
    ' worl' => 1,
    'world' => 1,
  }

On the other hand, this:

  my $text = "Hello, world";
  my $hash = ngram_counts({flankbreaks => 0}, $text, 5);

Produces the following ngrams:

  {
    'Hello' => 1,
    ' worl' => 1,
    'world' => 1,
  }

lowercase

If set to 0, casing is preserved. If set to 1, all letters are lowercased before counting ngrams. Default is 1.

    # Get all ngrams of size 4 preserving case
    $href_p = ngram_counts( {lowercase => 0}, $text, 4 );

punctuation

If set to 0 (default), punctuation is removed before calculating the ngrams. Set to 1 to preserve it.

    # Get all ngrams of size 2 preserving punctuation
    $href_p = ngram_counts( {punctuation => 1}, $text, 2 );

spaces

If set to 0 (default is 1), no ngrams containing spaces will be returned.

   # Get all ngrams of size 3 that do not contain spaces
   $href = ngram_counts( {spaces => 0}, $text, 3);

If you're going to request both types of ngrams, than the best way to avoid calculating the same thing twice is probably this:

    $href_with_spaces = ngram_counts($text[, $window]);
    $href_no_spaces = $href_with_spaces;
    for (keys %$href_no_spaces) { delete $href->{$_} if / / }

add_to_counts

This incrementally adds to the supplied hash; if $window is zero or undefined, then the window size is computed from the hash keys.

    add_to_counts($more_text, $window, $href)

TO DO

Look further into the tests. Sort them and add more.

AUTHOR

Maintained by Alberto Simoes, "[email protected]".

Previously maintained by Jose Castro, "[email protected]". Originally created by Simon Cozens, "[email protected]".

COPYRIGHT AND LICENSE

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