In an effort to be transparent with our roadmap, we publish important changes we are working on in advance. We will try to stick with our dates, but usually it is much easier to commit to month or quarter estimates for some of the harder changes. Since we are a small team and not all dependencies are in our control, dates can change and we will give our best to communicate updates as soon as possible. This roadmap intends to show future changes which may affect existing applications and also includes new features we are working on.
This is essentially collaborative filtering based on user profile similarity. Relevance will be calculated based on similarity between entities and terms users read about.
This feature will identify clusters of individuals geographically near a user’s area of interest, where “near” is defined as the neighborhood, hometown or country selected by the user.
This is something we wanted to implement for quite some time now, for two reasons:
- debugging and internal analysis
- being able to present to the user why they were served a piece of content
This can potentially allow us to provide end users with means to tune what is recommended to them (effectively, this means the different weights of the recommendation algorithm).