As a predictive analytics service, the impact of Youneeq’s recommendations is intimately tied to our ability to collect actionable data about our clients’ users. Interactive recommendations allow the end user to select what attributes an article should have and sends that information to the recommendation system to get new recommendations back. This process benefits the recommendation system in two ways: firstly, it allows the recommendation engines to get direct feedback to improve its recommendations to that specific user, and users that may be correlated to them; secondly, it provides information about a specific user and user group that is valuable outside the recommendation scope.
Any time an application is intending to gather information directly from an end user, it is important not to force any interaction upon the user — it should be an option not a requirement. If a person on a website feels he or she is being blocked by an application, that person is very likely going to avoid that application and even the website that it sits on. It is simple to have a small, unobtrusive options-type button that brings up a small menu of options for the user to choose from, which provides a large amount of utility for the user without disrupting that user’s experience. These options can offer improvement to pre-existing recommendation systems as well as add a possible additional revenue direction for the company using it.
