Spatial co-location pattern mining for location-based services in road networks

摘要: A new approach for mining spatial co-location patterns was presented.The neighborhood refined using network distances rather than Euclidean ones.An efficient algorithm to build the relationship graph proposed.The performance of proposed explored. With evolution geographic information capture and emergency volunteered information, it is getting more important extract knowledge automatically from large datasets. Spatial represent subsets features whose objects are often located in close proximity. Such pattern one most concepts context awareness location-based services (LBS). In literature, existing methods used events taking place a homogeneous isotropic space with distance expressed as Euclidean, while physical movement LBS usually constrained by road network. As result, interestingness value involving network-constrained cannot be accurately computed. this paper, we propose different method configurations geographical considered. First, define model linear referencing refine traditional ones. Then, considering that networks suffers expensive spatial-join operation, an way find all neighboring object pairs generating clique instances. By comparison previous approaches based on distance, can applied calculate probability occurrence Our experimental results real synthetic data sets show effective identifying which actually rely