/ 
Text
/ Text::Positional::Ngram

NAME

Text::Positional::Ngram

SYNOPSIS

This document provides a general introduction to the Text::Positional::Ngram module.

AUTHOR

Bridget Thomson McInnes, bthomson@d.umn.edu

BUGS

= head1 DESCRIPTION

1. Introduction

The Text::Positional::Ngram module is a module that allows for the retrieval of positional ngrams. An ngram is defined as a contiguous or non-contiguous sequence of 'n' tokens that occur within a window of at leaste 'n' tokens in the text. What constitutes as a 'token' can be defined by the user.

2. Ngrams

An ngram is a sequence of n tokens. The tokens in the ngrams are delimited by the diamond symbol, "<>". Therefore "to<>be<>" is a bigram whose tokens consist of "to" and "be". Similarly, "or<>not<>to<>" is a trigram whose tokens consist of "or", "not", and "to".

Given a piece of text, Ngrams are usually formed of contiguous tokens. For example, if we take the phrase:

to     be     or     not     to     be

The bigrams for this phrase would be:

to<>be<>     be<>or<>     or<>not<>

The trigrams for this phrase would be:

to<>be<>or<>     be<>or<>not<>     
or<>not<>to<>    not<>to<>be<>

3. Tokens

We define a token as a contiguous sequence of characters that match one of a set of regular expressions. These regular expressions may be user-provided, or, if not provided, are assumed to be the following two regular expressions:

\w+        -> this matches a contiguous sequence of alpha-numeric characters

[\.,;:\?!] -> this matches a single punctuation mark

For example, assume the following is a line of text:

"to be or not to be, that is the question!"

Then, using the above regular expressions, we get the following tokens:

to           be           or          not       
to           be           ,           that      
is           the          question    !

If we assume that the user provides the following regular expression:

[a-zA-Z]+  -> this matches a contiguous sequence of alphabetic characters

Then, we get the following tokens:

to           be           or          not       
to           be           that        is      
the          question

USAGE

use Text::Positional::Ngram

Text::Positional::Ngram Requirements

use Text::Positional::Ngram;

#  create an instance of Text::Positional::Ngram
my $text = Text::Positional::Ngram->new();

#  create the files needed and specify which
#  file you would like to get the ngrams from
$text->create_files("my_file.txt");

#  get the ngrams; the output will be stored in
#  'my_file.txt.ngram.<timestamp>'
$text->get_ngrams();

Text::Positional::Ngram Functions

1. create_files(@FILE)

Takes an array of files in which the ngrams are
to be obtained from. This function will creates the 
files that are required for the suffix array to be 
created. These files are defined as the name of the
first file entered in the FILE array and timestamped.

1. vocabulary file : converts tokens to integers prefix: 
2. snt file        : integer representation of corpus
3. sntngram file   : integer representation of the ngrams
                     and their frequency counts
4. ngram file      : ngrams and their frequencies

2. get_ngrams()

Obtains ngrams of size two and their frequencies
storing them in the given ngram file.

3. create_stop_list(FILE)

Removes n-grams containing at least one (in OR mode) 
stop word or all stop words (in AND mode). The default 
is OR mode. Each stop word should be a regular expression 
in this FILE and should be on a line of its own. These 
should be valid Perl regular expressions, which means that 
any occurrence of the forward slash '/' within the regular 
expression must be 'escaped'.

4. set_stop_mode(MODE)

OR mode removes n-grams containing at least 
one stop word and AND mode removes n-grams 
that consists of entirely of stop words. 
Default:  AND

5. set_token_file(FILE)

Each regular expression in this FILE should be on a line
of its own, and should be delimited by the forward slash 
'/'. These should be valid Perl regular expressions, which 
means that any occurrence of the forward slash '/' within 
the regular expression must be 'escaped'. 

NOTE: This function should be called before the 
function that creates the main files ie before 
create_files(FILE).

6. set_nontoken_file(FILE)

The set_nontoken_file function can be used when there 
are predictable sequences of characters that you know 
should not be included as tokens.

NOTE: This function should be called before the 
function that creates the main files ie before 
create_files(FILE).

7. set_remove()

Ignores Ignores n-grams that occur less than N times. 
Ignored n-grams are not counted and so do not affect 
counts and frequencies.

NOTE:  Should be set before you retrieve the ngrams, 
ie before you call the get_ngrams() function.

8. set_marginals()

The marginal frequencies consist of the frequencies of 
the individual tokens in their respective positions in
the n-gram.

NOTE:  Should be set before you retrieve the ngrams, 
ie before you call the get_ngrams() function.

9. set_new_line()

Prevents n-grams from spanning across the new-line
character
    
 NOTE: This function should be called before the 
function that creates the main files ie before 
create_files(FILE).

10. set_frequency(N)

Does not display n-grams that occur less than N times

NOTE:  Should be set before you retrieve the ngrams, 
ie before you call the get_ngrams() function.

11. set_window_size(N)

Sets the size of the window in which positional 
ngram can be found in.

12. set_ngram_size(N) Finds ngrams equal to size N Default : 2

NOTE:  Should be set before you retrieve the ngrams, 
ie before you call the get_ngrams() function.

13. set_ngram_file(FILE)

Prints the ngrams to FILE. If this is not set
the default is: <CORPUS FILE>.ngram.<timestamp>

NOTE: if an array of corpus files is defined,
the default is the first corpus file.

14. get_ngram_file()

Returns the name of the n-gram file.

15. get_ngram_count()

Returns the number of n-grams.

16. remove_files()

Removes the snt, sntngram and the vocab file.

COPYRIGHT

Copyright (C) 2004, Bridget Thomson McInnes

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.

Note: a copy of the GNU Free Documentation License is available on the web at http://www.gnu.org/copyleft/fdl.html and is included in this distribution as FDL.txt.

perl(1)