Friendica Communications Platform (please note that this is a clone of the repository at github, issues are handled there) https://friendi.ca
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 

503 lines
12 KiB

<?php
# Copyright (C) 2006-2010 Tobias Leupold <tobias.leupold@web.de>
#
# b8 - A Bayesian spam filter written in PHP 5
#
# This program is free software; you can redistribute it and/or modify it
# under the terms of the GNU Lesser General Public License as published by
# the Free Software Foundation in version 2.1 of the License.
#
# 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 Lesser General Public
# License for more details.
#
# You should have received a copy of the GNU Lesser 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.
/**
* Copyright (C) 2006-2010 Tobias Leupold <tobias.leupold@web.de>
*
* @license LGPL
* @access public
* @package b8
* @author Tobias Leupold
* @author Oliver Lillie (aka buggedcom) (original PHP 5 port)
*/
class b8
{
public $config = array(
'min_size' => 3,
'max_size' => 30,
'allow_numbers' => FALSE,
'lexer' => 'default',
'degenerator' => 'default',
'storage' => 'dba',
'use_relevant' => 15,
'min_dev' => 0.2,
'rob_s' => 0.3,
'rob_x' => 0.5
);
private $_lexer = NULL;
private $_database = NULL;
private $_token_data = NULL;
const SPAM = 'spam';
const HAM = 'ham';
const LEARN = 'learn';
const UNLEARN = 'unlearn';
const STARTUP_FAIL_DATABASE = 'STARTUP_FAIL_DATABASE';
const STARTUP_FAIL_LEXER = 'STARTUP_FAIL_LEXER';
const TRAINER_CATEGORY_FAIL = 'TRAINER_CATEGORY_FAIL';
/**
* Constructs b8
*
* @access public
* @return void
*/
function __construct($config = array(), $database_config)
{
# Validate config data
if(count($config) > 0) {
foreach ($config as $name=>$value) {
switch($name) {
case 'min_dev':
case 'rob_s':
case 'rob_x':
$this->config[$name] = (float) $value;
break;
case 'min_size':
case 'max_size':
case 'use_relevant':
$this->config[$name] = (int) $value;
break;
case 'allow_numbers':
$this->config[$name] = (bool) $value;
break;
case 'lexer':
$value = (string) strtolower($value);
$this->config[$name] = is_file(dirname(__FILE__) . DIRECTORY_SEPARATOR . 'lexer' . DIRECTORY_SEPARATOR . "lexer_" . $value . '.php') === TRUE ? $value : 'default';
break;
case 'storage':
$this->config[$name] = (string) $value;
break;
}
}
}
# Setup the database backend
# Get the basic storage class used by all backends
if($this->load_class('b8_storage_base', dirname(__FILE__) . DIRECTORY_SEPARATOR . 'storage' . DIRECTORY_SEPARATOR . 'storage_base.php') === FALSE)
return;
# Get the degenerator we need
if($this->load_class('b8_degenerator_' . $this->config['degenerator'], dirname(__FILE__) . DIRECTORY_SEPARATOR . 'degenerator' . DIRECTORY_SEPARATOR . 'degenerator_' . $this->config['degenerator'] . '.php') === FALSE)
return;
# Get the actual storage backend we need
if($this->load_class('b8_storage_' . $this->config['storage'], dirname(__FILE__) . DIRECTORY_SEPARATOR . 'storage' . DIRECTORY_SEPARATOR . 'storage_' . $this->config['storage'] . '.php') === FALSE)
return;
# Setup the backend
$class = 'b8_storage_' . $this->config['storage'];
$this->_database = new $class(
$database_config,
$this->config['degenerator'], date('ymd')
);
# Setup the lexer class
if($this->load_class('b8_lexer_' . $this->config['lexer'], dirname(__FILE__) . DIRECTORY_SEPARATOR . 'lexer' . DIRECTORY_SEPARATOR . 'lexer_' . $this->config['lexer'] . '.php') === FALSE)
return;
$class = 'b8_lexer_' . $this->config['lexer'];
$this->_lexer = new $class(
array(
'min_size' => $this->config['min_size'],
'max_size' => $this->config['max_size'],
'allow_numbers' => $this->config['allow_numbers']
)
);
}
/**
* Load a class file if a class has not been defined yet.
*
* @access public
* @return boolean Returns TRUE if everything is okay, otherwise FALSE.
*/
public function load_class($class_name, $class_file)
{
if(class_exists($class_name, FALSE) === FALSE) {
$included = require_once $class_file;
if($included === FALSE or class_exists($class_name, FALSE) === FALSE)
return FALSE;
}
return TRUE;
}
/**
* Validates the class has all it needs to work.
*
* @access public
* @return mixed Returns TRUE if everything is okay, otherwise an error code.
*/
public function validate()
{
if($this->_database === NULL)
return self::STARTUP_FAIL_DATABASE;
# Connect the database backend if we aren't connected yet
elseif($this->_database->connected === FALSE) {
$connection = $this->_database->connect();
if($connection !== TRUE)
return $connection;
}
if($this->_lexer === NULL)
return self::STARTUP_FAIL_LEXER;
return TRUE;
}
/**
* Classifies a text
*
* @access public
* @package default
* @param string $text
* @return float The rating between 0 (ham) and 1 (spam)
*/
public function classify($text)
{
# Validate the startup
$started_up = $this->validate();
if($started_up !== TRUE)
return $started_up;
# Get the internal database variables, containing the number of ham and
# spam texts so the spam probability can be calculated in relation to them
$internals = $this->_database->get_internals();
# Calculate the spamminess of all tokens
# Get all tokens we want to rate
$tokens = $this->_lexer->get_tokens($text);
# Check if the lexer failed
# (if so, $tokens will be a lexer error code, if not, $tokens will be an array)
if(!is_array($tokens))
return $tokens;
# Fetch all availible data for the token set from the database
$this->_token_data = $this->_database->get(array_keys($tokens));
# Calculate the spamminess and importance for each token (or a degenerated form of it)
$word_count = array();
$rating = array();
$importance = array();
foreach($tokens as $word => $count) {
$word_count[$word] = $count;
# Although we only call this function only here ... let's do the
# calculation stuff in a function to make this a bit less confusing ;-)
$rating[$word] = $this->_get_probability($word, $internals['texts_ham'], $internals['texts_spam']);
$importance[$word] = abs(0.5 - $rating[$word]);
}
# Order by importance
arsort($importance);
reset($importance);
# Get the most interesting tokens (use all if we have less than the given number)
$relevant = array();
for($i = 0; $i < $this->config['use_relevant']; $i++) {
if($tmp = each($importance)) {
# Important tokens remain
# If the token's rating is relevant enough, use it
if(abs(0.5 - $rating[$tmp['key']]) > $this->config['min_dev']) {
# Tokens that appear more than once also count more than once
for($x = 0, $l = $word_count[$tmp['key']]; $x < $l; $x++)
array_push($relevant, $rating[$tmp['key']]);
}
}
else {
# We have less than words to use, so we already
# use what we have and can break here
break;
}
}
# Calculate the spamminess of the text (thanks to Mr. Robinson ;-)
# We set both hamminess and Spamminess to 1 for the first multiplying
$hamminess = 1;
$spamminess = 1;
# Consider all relevant ratings
foreach($relevant as $value) {
$hamminess *= (1.0 - $value);
$spamminess *= $value;
}
# If no token was good for calculation, we really don't know how
# to rate this text; so we assume a spam and ham probability of 0.5
if($hamminess === 1 and $spamminess === 1) {
$hamminess = 0.5;
$spamminess = 0.5;
$n = 1;
}
else {
# Get the number of relevant ratings
$n = count($relevant);
}
# Calculate the combined rating
# The actual hamminess and spamminess
$hamminess = 1 - pow($hamminess, (1 / $n));
$spamminess = 1 - pow($spamminess, (1 / $n));
# Calculate the combined indicator
$probability = ($hamminess - $spamminess) / ($hamminess + $spamminess);
# We want a value between 0 and 1, not between -1 and +1, so ...
$probability = (1 + $probability) / 2;
# Alea iacta est
return $probability;
}
/**
* Calculate the spamminess of a single token also considering "degenerated" versions
*
* @access private
* @param string $word
* @param string $texts_ham
* @param string $texts_spam
* @return void
*/
private function _get_probability($word, $texts_ham, $texts_spam)
{
# Let's see what we have!
if(isset($this->_token_data['tokens'][$word]) === TRUE) {
# The token was in the database, so we can use it's data as-is
# and calculate the spamminess of this token directly
return $this->_calc_probability($this->_token_data['tokens'][$word], $texts_ham, $texts_spam);
}
# Damn. The token was not found, so do we have at least similar words?
if(isset($this->_token_data['degenerates'][$word]) === TRUE) {
# We found similar words, so calculate the spamminess for each one
# and choose the most important one for the further calculation
# The default rating is 0.5 simply saying nothing
$rating = 0.5;
foreach($this->_token_data['degenerates'][$word] as $degenerate => $count) {
# Calculate the rating of the current degenerated token
$rating_tmp = $this->_calc_probability($count, $texts_ham, $texts_spam);
# Is it more important than the rating of another degenerated version?
if(abs(0.5 - $rating_tmp) > abs(0.5 - $rating))
$rating = $rating_tmp;
}
return $rating;
}
else {
# The token is really unknown, so choose the default rating
# for completely unknown tokens. This strips down to the
# robX parameter so we can cheap out the freaky math ;-)
return $this->config['rob_x'];
}
}
/**
* Do the actual spamminess calculation of a single token
*
* @access private
* @param array $data
* @param string $texts_ham
* @param string $texts_spam
* @return void
*/
private function _calc_probability($data, $texts_ham, $texts_spam)
{
# Calculate the basic probability by Mr. Graham
# But: consider the number of ham and spam texts saved instead of the
# number of entries where the token appeared to calculate a relative
# spamminess because we count tokens appearing multiple times not just
# once but as often as they appear in the learned texts
$rel_ham = $data['count_ham'];
$rel_spam = $data['count_spam'];
if($texts_ham > 0)
$rel_ham = $data['count_ham'] / $texts_ham;
if($texts_spam > 0)
$rel_spam = $data['count_spam'] / $texts_spam;
$rating = $rel_spam / ($rel_ham + $rel_spam);
# Calculate the better probability proposed by Mr. Robinson
$all = $data['count_ham'] + $data['count_spam'];
return (($this->config['rob_s'] * $this->config['rob_x']) + ($all * $rating)) / ($this->config['rob_s'] + $all);
}
/**
* Check the validity of the category of a request
*
* @access private
* @param string $category
* @return void
*/
private function _check_category($category)
{
return $category === self::HAM or $category === self::SPAM;
}
/**
* Learn a reference text
*
* @access public
* @param string $text
* @param const $category Either b8::SPAM or b8::HAM
* @return void
*/
public function learn($text, $category)
{
return $this->_process_text($text, $category, self::LEARN);
}
/**
* Unlearn a reference text
*
* @access public
* @param string $text
* @param const $category Either b8::SPAM or b8::HAM
* @return void
*/
public function unlearn($text, $category)
{
return $this->_process_text($text, $category, self::UNLEARN);
}
/**
* Does the actual interaction with the storage backend for learning or unlearning texts
*
* @access private
* @param string $text
* @param const $category Either b8::SPAM or b8::HAM
* @param const $action Either b8::LEARN or b8::UNLEARN
* @return void
*/
private function _process_text($text, $category, $action)
{
# Validate the startup
$started_up = $this->validate();
if($started_up !== TRUE)
return $started_up;
# Look if the request is okay
if($this->_check_category($category) === FALSE)
return self::TRAINER_CATEGORY_FAIL;
# Get all tokens from $text
$tokens = $this->_lexer->get_tokens($text);
# Check if the lexer failed
# (if so, $tokens will be a lexer error code, if not, $tokens will be an array)
if(!is_array($tokens))
return $tokens;
# Pass the tokens and what to do with it to the storage backend
return $this->_database->process_text($tokens, $category, $action);
}
}
?>