Continuous Probabilistic Skyline Queries On Moving Objects With Uncertain Location

Skyline query aims at finding out the objects in the database those attribute vectors are notdominated by any other objects. Skyline query is of great importance, such as multi-criteriadecision making, preference queries, data mining and data visualization. Since it was proposed,skyline computation has received considerable attention in the database community.Early studies for skyline operation usually assume that static data objects in the centralizeddata base. However, the objects are actually dynamic in the real world. As the development ofwireless communication systems, devices become smaller and smaller, as well as onboardsystems became commonplace whose position information is changing as time goes on. Somescholars did research on continues skyline queries for moving objects. Uncertainty is anotherinherent nature of the data. In this paper we assume that the objects are moving with uncertaintylocation. Computing a probabilistic skyline is much more complicated than computing a skylineon certain data. Due to the uncertainty location we couldn’t sure the dominant relationshipbetween two moving objects. A reasonable probability-based skyline model should be built toresolve this problem. In this paper, continuous probabilistic skyline queries for moving objectswith uncertainty under unconstrained space and road network environment are studied. Themain work can be described briefly as follows:1) Through the analysis and comparison of existing uncertainty models, an uncertainregional model was established, which is an unconstrained space environment. Wegave the definition of dominant probability and skyline probability. We did a depthanalysis about the distance function between moving objects how to affect thedominant probability in moves objects. Events affecting p-skyline are defined toupdate p-skyline set continuously. Based on above analysis, we proposed analgorithm to address the continuous probabilistic skyline queries on moving objectin unconstrained space environment—(Event triggered Continuous ProbabilisticSkyline query for uncertain moving object, U-ECPS).2) We choose a segment model for the road network environment. The dominantprobability and skyline probability are defined. Through the analysis andcomparison of the network distance between two moving objects the eventsaffecting p-skyline are defined. And then we proposed an algorithm to address the continuous probabilistic skyline queries on moving object in road network—(Continuous Probabilistic Skyline query for Uncertain moving object in Roadnetwork, PSUR).3) The algorithms we’ve proposed do not traverse the whole database to return theanswers. They update the answers through track and process the events. With theresult the efficiency of the algorithms is increased. Extensive experiments havedemonstrated the efficiency and effectiveness of our proposed algorithms.