Skip to main content
Skip table of contents

Direct Integration with Ncanto

The recommended way to use Ncanto for recommendations is to use Recommendations Façade, as explained in Recommendations Façade - Ncanto integration.

This section is for operators that started using Ncanto before Recommendations Façade was available.

As well as having a Recommendations Façade adapter,  OpenTV Platform 3.x is directly integrated with Ncanto from XroadMedia to enable you to provide personalised recommendations to your users.

OpenTV Platform provides the ability to load your media assets into the Ncanto platform, to ensure that user interactions with the content via the OpenTV Platform are used to provide meaningful subsequent recommendations. The Platform manages the loading and subsequent maintenance of the content (add, update, delete, and expire) in NCanto. 

A client application needs to do two things (interacting with Content Discovery Gateway (CDG) in both cases):

  1. Report user activity (what the user watches, what they download, what they record, etc.) to provide Ncanto with the data it needs to generate personalised recommendations.
  2. Request recommendations and display them to the user.

Ncanto has a concept of "contexts". A context provides the most relevant recommendations for a specific scenario of use. A context is a combination of configuration parameters (for example, user ratings, content age (production year), how much longer the content will be valid, time of day, and so on) and weightings for these parameters.

You need to be aware of the available contexts so that your application can request the right recommendations. (CDG has an API to get the available contexts – see  Content Discovery Gateway (CDG) API documentation.)

OPF is deployed with a default set of contexts, for example, "Because you watched", "Latest arrivals", and so on. You can modify these contexts and create your own using XroadMedia's Piloto application – see  Editing contexts.

Supported use cases and additional information


JavaScript errors detected

Please note, these errors can depend on your browser setup.

If this problem persists, please contact our support.