Mining Maximal Dynamic Spatial Colocation Patterns

A spatial co-location pattern represents a subset of features whose instances are prevalently located together in geographic space. Although many algorithms mining have been proposed, there still some problems: 1) they miss meaningful patterns (e.g., {Ganoderma_lucidumnew, maple_treedead} and {water_hyacinthnew(increase), algaedead(decrease)}), get the wrong conclusion that two or more increase/decrease (i.e., new/dead) same/approximate proportion, which has no effect on prevalent patterns. 2) Since number is very large, efficiency existing methods low to mine Therefore, first, we propose concept dynamic can reflect relationships among features. Second, small maximal derive all patterns, improve obtaining Third, an algorithm for pruning strategies. Finally, effectiveness method proposed as well strategies verified by extensive experiments over real/synthetic datasets.