Effective reverse K-nearest neighbor query based on revised R∗-tree in spatial databases

This paper presents a novel algorithm for reverse k nearest neighbor queries (RkNN), based on the Revived R*-tree index structure. Existing incremental methods for RkNN have the flowing drawbacks: (i) they cannot support objects in multidimensional space, (ii) their methods are low efficient for incremental query. To solve such RkNN problem efficiently, we propose a novel incremental RkNN algorithm, applied to multidimensional spatial databases. In this algorithm, we introduce a counter for every entry of RR*-tree index structure, which marks the number of nearest neighbor and thus offers the information about the influences of a query point. Experiments analyze synthetic and real data sets and show that our solution is more efficient traditional reverse nearest neighbor queries.