Personalize your news feeds
We help publishers around the globe create a high-quality metadata foundation and build innovative media products.
Our Personalization & Recommendation engine can be used to build a wide array of experiences to cater to all kinds of users.
Personalization & Recommendation engine
We want to help our clients with excellent ways of distributing news content. Our Personalization & Recommendation engine is built for the news industry and can be customized to fit your specific organization and goals.
A combination of six pillars
Some articles are relevant for a few hours, while other articles are relevant for days or even weeks after publication.
The freshness of an article is a foundational aspect for serving relevant content.
Often there is a strong geographical context for a news story, and the story is naturally more relevant to people that live nearby.
To cater to this, we have features to allow articles that are close to your user rank highly in the recommendations.
Interests & personalization
Cultural closeness, or in simple terms, relatability is a foundational aspect of news valuation. If you can personally relate to a piece of content, you are more likely to perceive it as interesting.
Therefore we have chosen to focus a lot on personalization efforts, working with both interpreted interests based on read history and explicit interests as part of user account settings.
The article performance measures how many users have read an article during a time. This is the measure that identifies trending content. Article performance is a simple and potent feature that can influence the recommendations.
Editorial news value
News editors often deem certain articles to be more important and of interest to the general public. This is sometimes indicated by an article’s news value set by an editor. Our recommendation service takes into account the editorial input for news value.
Context as a strong signal
Recommendations are often presented adjacent to an article, and in this case, this article can be used as a context to guide the recommendation engine to content that the user is interested in right now.