Batch processing of top-K spatial-textual queries

Top-k spatial-textual queries have received significant attention in the research community. Several techniques to efficiently process this class of queries are now widely used in a variety of applications. However, the problem of how best to process multiple queries efficiently is not well understood. Applications relying on processing continuous streams of queries, and offline pre-processing of other queries could benefit from solutions to this problem. In this work, we study practical solutions to efficiently process a set of top-k spatial-textual queries. We propose an efficient best-first algorithm for the batch processing of top-k spatial-textual queries that promotes shared processing and reduced I/O in each query batch. By grouping similar queries and processing them simultaneously, we are able to demonstrate significant performance gains using publicly available datasets.