| Summary: | The continuous queries over spatial-textual data streams (CQST) are widely used in location-based services, which continuously monitor the results satisfying spatial and textual constraints over updated data streams. To match the objects over the data streams with CQST as soon as possible, building efficient filtering techniques on the CQST is the key. The approach to evaluating CQST is to select the appropriate spatial and textual index to organize the CQST, construct efficient filtering strategies to improve the spatial and textual filtering capabilities of the index, filter large number of unpromising queries for the objects over the data streams, avoid costly verification cost and improve the efficiency of matching objects and queries. The existing works construct spatial-textual hybrid index by using limited spatial index and textual index, and the difference in evaluation performance depends on the filtering strategies used, i.e. the techniques of improving the filtering performance of the index. This paper takes the techniques of evaluating CQST as the research object, introduces the framework and challenges of evaluating CQST, reviews and compares the spatial and textual filtering techniques of evaluating CQST on a central server and distributed clusters, including the adopted spatial-textual hybrid index, the spatial and textual filtering strategies and the combination scheme of the spatial and textual index to improve the filtering performance of the index, analyzes and summarizes their advantages and disadvantages, and discusses possible future research directions.
|
|---|
