> The challenge is to generate a state machine in which the weights are dynamically adjusted as the user navigates the graph, resulting in the optimal direction. The best we can do is to analyze the input of other users to generate weights. "Crowd-sourced" weights.
I think you are right on point 1, but point 2, perhaps we can generate the weights for te individual user, by asking a (very) few simple 'onboarding questions' that help generate the hierarchy that would suit the user?
You sometimes see sites reaching for this kind of approach with the 'For user type X', 'for user type Y' top-level navigation - but this then just leads to multiple static heirarchies.
I think you are right on point 1, but point 2, perhaps we can generate the weights for te individual user, by asking a (very) few simple 'onboarding questions' that help generate the hierarchy that would suit the user?
You sometimes see sites reaching for this kind of approach with the 'For user type X', 'for user type Y' top-level navigation - but this then just leads to multiple static heirarchies.