A Naive Bayesian classification library for PHP with support for different tokenizers to optimize string classification.

• Introduction
• Examples
• Running Tests
• Troubleshooting
• Changelog

Naive Bayesian Classification provides a computationally cheap, but reasonably accurate method of determining which "group" a provided string belongs to given a prior training set.

PHP Classifier supports any number of groups to classify strings against and uses logarithmic counting to prevent integer underflow when using larger sets.

$tokenizer = new HybridLogic\Classifier\Basic;
$classifier = new HybridLogic\Classifier($tokenizer);

$classifier->train('hot', 'The sun is hot');
$classifier->train('hot', 'It was a warm day in the sun');
$classifier->train('hot', 'This tea is hot!');

$classifier->train('cold', 'This ice is very cold!');
$classifier->train('cold', 'It\'s cold at night');
$classifier->train('cold', 'Ice formed on my at over night');

$groups = $classifier->classify('It was chilly last night');

phpunit tests

@todo

• [2013-01-10] Initial port from KoBayes