At Morpht, our mission is to innovate in content delivery to foster user engagement. We want to make our websites smarter for a better user experience. Our recent proof-of-concept project, GovFlix, demonstrates the potential of content recommendations to deliver more targeted content to users based on their browsing behaviour. Built on a Drupal CMS site using the GovCMS SaaS platform, this project exemplifies how recommendation systems can transform content discovery.
As content designers and creators we are familiar with the concepts of categorising content and managing relationships between items. Traditionally the web has been arranged around the concepts of human curated content where knowledgeable editors and domain experts manage content and design the user experience. A common pattern is to define “related” links to keep the user journey flowing once the main content has been consumed.
Developments in machine learning have given rise to a new era of content management where content can be organised according to (i) human behaviour where the interests of the user can be derived and (ii) semantic analysis of text where concepts can be extracted for meaning. These two advancements open the way for new ways to be developed for users to navigate content.
These advancements are being developed as SaaS products where content and behaviour can be combined to provide dynamic content recommendations to users based on their browsing history and their interests. the wisdom of the crowd can be used to provide recommendations which may not be otherwise made by an editor or domain expert.
In this presentation, we cover these advancements and take a look at how Recombee, a recommender as a service, has been integrated into a GovCMS SaaS website without the need for any new modules. This easy integration path opens the way for government agencies to easily experiment with recommendation systems without the need for large investments in technology or implementation.
The concept of GovFlix
GovFlix is a curated video content site focusing on the government technology sector. It showcases how features like “top picks,” “trending now,” and “related content” can offer personalised experiences, making content easier and more engaging to navigate.
Innovative content publishing
GovFlix represents one possible future of content publishing with personalised user experiences. The interface is able to adapt so that each user is presented with a unique experience based on what they are interested in. This approach isn't limited to media; it can extend to education, publishing, and other government services, improving content relevance and user satisfaction.
The way content has been displayed has been reimagined. Rather than using traditional Views which list out content by date, the listing blocks are now generated from a recommendation engine. The aim has been to personalise all the content listings so that users are presented with the content most relevant to them. This kind of approach is suitable for sites where users have a strong intent and would like to be presented with similar content.
Recombee integration
Recombee served as the recommendation service, managing the delivery of personalised content. Setting up Recombee involved defining a schema for the Recombee index, injecting a data feed from Drupal and finally defining a number of Scenarios for recommendations. Recombee delivers recommendations via a JSON endpoint which can then be consumed and transformed client side in Drupal to deliver personalised HTML to users.
Why GovFlix matters
The presentation will illustrate how a content recommender system can revolutionise the user experience for users across a wide range of industries. With AI-driven technologies becoming more accessible, government agencies can now explore adopting these innovations to enhance digital experiences for citizens.
