Spatial Data Mining: Database Primitives, Algorithms and Efficient DBMS Support

摘要: Spatial data mining algorithms heavily depend on the efficient processing of neighborhood relations since neighbors many objects have to be investigated in a single run typical algorithm. Therefore, providing general concepts for as well an implementation these will allow tight integration spatial with database management system. This speed up both, development and execution algorithms. In this paper, we define graphs paths small set primitives their manipulation. We show that are supported by proposed basic operations. For finding significant patterns, only certain classes “leading away” from starting object relevant. discuss filters allowing such which significantly reduce search space Furthermore, introduce indices our primitives. implemented top commercial The effectiveness efficiency approach was evaluated using analytical cost model extensive experimental study geographic database.