Naxrun's Website Content Recommendation engine is free for anyone to use. We have developed it as a proof of concept, that illustrates how a website can display the most relevant content, based on a visitor's previous activity.
The current version is v0.1 beta.
The engine is split into 2 main scripts:
- logger.php
- recommendation.php
The logger.php script is used to log every site interaction a site visitor has. It takes the following parameters:
- v_ip - The visitor's IP address (only IPv4 supported)
- v_ref - The visitor's referrer URL
- v_uid - Any unique visitor ID - eg. a fingerprint or ID stored in a cookie
- p_cat - Page category ID, which must be an integer
- p_url - Page URL
The logger.php script logs all these values in a MySQL database. All parameter values are submitted using POST.
The script will return a JSON element with info about the inserted data and a HTTP status code. Please note, that providing no input will cause the script to insert NULL values for everything, except 'p_cat', which will be set to '1'.
The other script - recommendation.php - provides a page category ID recommendation, based on either the visitor's IP address or the unique ID attached to the visitor. The script takes these parameters:
- v_ip - The visitor's IP address (only IPv4 supported)
- v_uid - Any unique visitor ID - eg. a fingerprint or ID stored in a cookie
All parameter values are submitted using POST.
The script will return a JSON element with recommendations based on the visitor's IP and/or unique ID, and a HTTP status code.
The only thing needed to run this is PHP 5.6 and MySQL 5.5.
Start by setting up the database. Use the 'setup.sql' to setup the database.
Download this repository to your web server. Replace the values in 'includes/c_db.php' with your database info.
As mentioned, this is merely a proof of concept, so we haven't built any standardized way to interact with the scripts. You can use them both server-side and client-side. It's all up to you.