Scalable Spatial Predictive Query Processing for Moving Objects

Abstract

A fundamental category of location based services relies on predictive queries which consider the anticipated future locations of users. Predictive queries at- tracted the researchers' attention as they are widely used in several applications including traffic management, routing, location-based advertising, and ride shar- ing. This thesis aims to present a generic and scalable system for predictive query processing on moving objects, e.g., vehicles. Inside the proposed system, two frameworks are provided to work on two different environments, (1) Panda framework for Euclidean space, and (2) iRoad framework for road network. In- side the iRoad system, a novel data structure named Predictive Tree (P-Tree) is proposed to index the anticipated future locations of objects on road networks. Unlike previous work in supporting predictive queries, the target of the proposed system is to: (a) support long-term query prediction as well as short term predic- tion, (b) scale up to large number of moving objects, and (c) efficiently support different types of predictive queries, e.g., predictive range, KNN, and aggregate queries.