The Farthest Spatial Skyline Queries

Pareto-optimal objects are favored as each of such objects has at least one competitive edge against all other objects, or “not dominated.” Recently, in the database literature, skyline queries have gained attention as an effective way to identify such pareto-optimal objects. In particular, this paper studies the pareto-optimal objects in perspective of facility or business locations. More specifically, given data points P and query points Q in two-dimensional space, our goal is to retrieve data points that are farther from at least one query point than all the other data points. Such queries are helpful in identifying spatial locations far away from undesirable locations, e.g., unpleasant facilities or business competitors. To solve this problem, we first study a baseline algorithm TFSS and propose an efficient progressive algorithm BBFS, which significantly outperforms TFSS by exploiting spatial locality. We also develop an efficient approximation algorithm to trade accuracy for efficiency. We validate our proposed algorithms using extensive evaluations over synthetic and real datasets.