News Classification project

Overview

• This is a machine learning project for classifying news article, paragraph, text into 5 categories: business, entertainment, politics, sport and technology.

• The training dataset are 2225 bbc news articles already labeled into 5 classes (business, entertainment, politics, sport and technology) from http://mlg.ucd.ie/datasets/bbc.html.

• The model for classification is Support Vector Machine with N-gram for vectorizing words, sentences using Php-ml library.

• The accuracy is 0.956.

• Languages: PHP, Html, JQuery

Installation

• Pull this example: git clone https://github.com/cuongnguyenngoc/News-Classification.git
• Then cd News-Classification
• Then run composer install to install all the requirements
• Now you can move this directory News-Classification to xampp/htdocs folder to run the php code (remeber to start apache server).

Learn more about php-ml here https://github.com/php-ai/php-ml-examples