WP Recommender

This plugin finds and suggests relevant, potentially interesting content from
your blog to visitors. The goal is to push the relevant content to the visitor
with the prospect of increase of page views on your blog and eventually
satisfied and returning visitor.

Plugin is able to recommend similar posts and pages using two different
approaches. The first is content based and produces list of the most similar
items based on words occurring in both items. Advantage of this approach lies
in is its simplicity for users to install and it works straight from the start
(there’s no so-called “cold start” period from the collaborative filtering
approach).

The second approach is employing collaborative filtering to identify post and
pages that might interest user. Suggested items are calculated from the users’
browsing history that plugin logs on the aliiike server. Server accepts such
logs only if you are registered and you created an account. Disadvantage of
this approach is the fact that you need to wait until enough data is logged.
If the traffic on your site is low it can last for a while. Advantage of this
approach is the fact that you are modeling users’ behaviour (rather than the
contents), which in normal circumstances should easily outperform the content
based approach.

Important aspect of plugin is its ability to serve recommendation lists to a
settable percentage of users only. In combination with the Google Analytics
this feature allows you to perform so-called A/B testing to measure the
influence of the recommended items on your visitors’ browsing behavior (or
sales increase or decrease in case you are running a web shop).

See the Aliiike recommender homepage
as well as the rest of the site for more information.