Fast Matching of Detour Routes and Service Areas

Siqiang Luo,Reynold Cheng,Xiaokui Xiao,Ben Kao,Shuigeng Zhou,Jiafeng Hu

2016-01-01

Abstract:Novel road-network services, such as ride sharing, spatial crowdsourcing and keyword-aware routing often need to find out whether there exists a route that can go through (or match) a service area. A ride-sharing system, for instance, may need to find for a moving taxi a route that passes through a region in which a passenger is located. A fundamental question is: given two locations s and t, and an area A on the road network, is there a “reasonably short” route from s to t that also crosses A? This Route and Area Matching (ROAM) query is important as 1) its efficient evaluation allows the application involved to recommend appropriate services to objects in an online manner; 2) its algorithm can be incorporated into some sophisticated systems to optimize their performance. Despite its importance, the ROAM query has not been well studied especially when “service area” instead of “service point” is concerned. We examine efficient ROAM query algorithms. Particularly, we develop solutions for testing the existence of a ρ-route, which is an s-t path that passes through A, with a length of at most (1 + ρ) times the shortest distance between s and t. We propose index-free and index-based algorithms. Our index structures, with a size linear to the number of road network nodes, enable significant query improvement over baseline approaches. We have performed extensive experiments on several large road network datasets to validate the efficiency and scalability of our approaches. In addition, we incorporate our efficient algorithm to an existing keyword-aware routing system and one dynamic ride-sharing system. Our experiments show that an efficient ROAM query algorithm helps improve the performance of these systems.