Topic

Recommendations

Recommendations deliver content that is fresh, popular, related and customised to user interests and preferences.

Recent and similar content

Displays content a user recently interacted with and supplements it with suggestions of content similar to what they’re interested in.
Collaborative filtering

Analyses user behaviour and recommends content similar to what other users with similar tastes have enjoyed.
Latest and popular

Displays recently released content and supports it with trending content that's garnered the most engagement.
Customisation

The system can be customised to specific business needs and goals, including different recommendation types and layouts.

Powered by Recombee

Morpht has partnered with Recombee, a leading AI-powered Recommender-as-a-Service, integrated via a purpose-built Drupal module.

Track real interactions

The system tracks how visitors interact with your site content as they browse.

Apply the wisdom of the crowd

Collaborative filtering analyses that behaviour across all users to surface meaningful patterns.

Surface it automatically

Relevant recommendations appear in defined sections of your site, the same approach used by Amazon, Netflix and Spotify.

What is Recombee and how does it work with Convivial for GovCMS?

Recombee is a leading AI-powered Recommender-as-a-Service platform. Morpht has built a Drupal module that connects Convivial for GovCMS to Recombee, tracking user interactions with your content and using collaborative filtering, similar to the recommendation engines behind Amazon, Netflix and Spotify, to surface relevant suggestions in defined sections of your site.

What kind of recommendations can I show on my site?

Convivial for GovCMS supports several recommendation types: recent and similar content based on what a visitor has already viewed, collaborative filtering based on what similar users have engaged with, and latest and popular content based on trending engagement. Recommendation types and layouts can be customised to suit specific business goals.

Does Recommendations require personalisation to already be set up?

Recommendations can be used independently. It draws on its own user interaction tracking through Recombee rather than relying on the Convivial Profiler data used for on-page personalisation, though the two features work well together.

How much user activity is needed before recommendations improve?

Recombee's algorithms improve as more interaction data is collected, since recommendations are based on the "wisdom of the crowd" across your site's visitors. Even with a modest amount of traffic, the system can surface recent and popular content while collaborative filtering strengthens over time.

How is Recommendations different from Personalisation?

Recommendations surfaces content a visitor hasn't seen yet, based on the collective behaviour of all users on your site. Personalisation instead tailors the tasks, promotions and content on the page a visitor is already viewing, based on their own individual profile. The two are complementary rather than overlapping.