An interactive approach to route search
Yaron Kanza, Roy Levin, Eliyahu Safra, Yehoshua Sagiv · 2009
In a probabilistic route search, there is a start location, a target location, and search queries Q1, ..., Qn. Each Qi has an answer set Ai consisting of geo-spatial objects and their probabilities. The probability of an object o ∈ Ai specifies the likelihood that o satisfies Qi. The goal is to compute a route that is short and yet has a high probability of satisfying all the Qi. This paper investigates interactive route search. Upon arrival at each object, the user provides feedback specifying whether the object satisfies its corresponding query. The goal is to compute the next object to be visited, based on the feedback. Several heuristic algorithms are given and compared experimentally.