13. Decision Paths
By showing how decision paths work (for example visualizing the layers in a recommender system) users can get an idea about why they get recommended certain things. The decision paths can be made adjustable by the user to tune their recommendations.
What can designers do with this pattern?
Example of peer pickers
Still to be found in the wild.
Input or output?
How do users control the algorithm?
Recommendation paths in itself do not offer the user control. Imaginable is that users can change paths. As has been implemented in some experimental systems.
How do users understand the control?
Users can gain a notion of how recommendations are composed.
Some examples of use
Please let us know if you have a related reference with the scientific research.
- Examples can be found in He et al. (2015)
Comments & Talks
React on this algoritmic affordance and get in touch with the publisher over this algoritmic affordance Koen van Turnhout.