About Content Recommendations (CR)

This is for:

In this article

Content recommendations example

Coveo Machine Learning (Coveo ML) Content Recommendation (CR) models recommend the most relevant content for each user, by learning from clicks and view events across all users' previous sessions. These recommendations can be easily deployed in any web page, community, marketplace, or other digital property.

In more detail, the Coveo ML CR algorithm is based on the repeated occurrence of clicks and view events within a user visit. When two of these events commonly occur within multiple sessions, the algorithm learns that they’re inherently linked. When one event is seen, the model recommends the other. From the user’s perspective, the recommended event is in fact the link to the recommended item.

The recommendations can be interpreted as "People who viewed this page also viewed the following pages."

Members with the required privileges can create, manage, and deploy CR models.

Note

CR model suggestions are only based on the user language, since view events aren’t logged from a search hub.

Content recommendations example

Your public support portal has a large inventory of content (blog posts, support articles, videos, community posts) that may provide solutions to issues customers are facing with your products. The portal is configured to send all view events to Coveo Usage Analytics (Coveo UA).

To enhance the user experience, the support portal incorporates the following recommendation components:

A Videos we Recommend for you carousel that filters recommended content to only display items of the video content type.
A People Like You Also Viewed component that compares the current user actions history to those of similar users to recommend content that hasn’t yet been seen by the current user, but has been useful to other similar users.
A recommendation component that leverages custom context to provide personalized recommendations based on the user’s attributes.