Research On Approaches Of Complex Queries In Spatio-temporal Databases

Advances in mobile computing, wireless communication and wide deployment of positioning devices have made the applications of spatio-temporal databases become wider and deeper. Most of these applications need spatio-temporal databases to efficiently process various queries. The query efficiency is an important benchmark of the performance of spatio-temporal databases. The complexity of spatiotemporal data and the introduction of new affect factors make the spatio-temporal queries more complicated, which are not alike the previous simple queries that only consider single query object and position factor. It is urgent to give out query processing approaches to efficiently process these complicated spatio-temporal queries. Therefore, how to provide all kinds of efficient query processing approaches for spatiotemporal objects is one of research hotspots in the reseach spectrum of spatio-temporal databases currently.Nearest keyword aggregate query introduces the keyword search into aggregate queries. The spatial objects contain location and keyword information that is represented by a set of keywords. The distance to nearest keyword is defined to be the distance between query point and its nearest object which contains the keyword. Nearest keyword aggregate distance is the aggregate value on the multiple nearest keyword distances. Nearest keyword aggregate query search a query point with the minimum nearest keyword aggregate distance from a set of query points. Nearest keyword aggregate query enriches the types of spatial keyword queries. Utilizing the minimum nearest keyword distance can efficiently filter data objects and speed up the calculation of the nearest keyword distance. The unqualified query points can be directly deleted through the minimum nearest keyword aggregate distance. The numbers of node accesses on query points and data objects are both reduced.Objects commomly have a great impact on its nearest neighbors, the number of its reverse nearest neighbor indicates its influence. Continuous Reverse Nearest Aggregate Query continuously returns a query point with the most influence from a set of query points. As the extension of reverse nearest aggregate queries on moving objects, Continuous Reverse Nearest Aggregate Query focused on how to efficiently monitor the changes of results in a short period of time. Utilizing the minimum and maximum number of query points which are closer to a cell than a certain query point can fasten the speed of identification of reverse nearest neighbors. The search region of reverse nearest neighbors can shrink by the use of minimum number, and all moving objects in a cell can be directly identified to be true reverse nearest neighbors by maximum number. Identifing the query points to be monitored by the number of reverse nearest neighbor candidates reduces the candidates to be checked further, so it can greatly reduce the reponse to query.The actual distance between two points can be affected by the obstacles in real world The obstructed distance between two points in an obstructed space is the length of shortest path without cutting through any obstacle. Continuous Obstructed Range Query considers the constraint of obstacles in continuous range query, it continuously all moving objects the obstructed distances of which are in the range of a positive number. How to simplify the obstructed distance calculations and reduce the number of obstructed distance calculations becomes the key of query approaches. The obstructed distance calculations are simplified by dividing a big visiblilty graph into multiple small local visibility graphs. Utilizing the relative position between moving objects and obstacles can quickly identify whether the moving objects are in the obstructed range, which results in the reduction of the number of obstructed distance calculations. The computed visilbility graph can be used to speed up the obstructed distance calculations, therefore the efficiency of query approaches is improved.