Incremental Maintenance of Discovered Spatial Colocation Patterns

Unlike the traditional incremental updating problem for discrete data, the appended data to spatial dataset may introduce lots of new relations between the added events and the existing events. Moreover, as the measure in mining of colocation patterns, participation index is complicated to handle compared with simply support counter. Thus, the incremental maintenance of colocation patterns for dynamic spatial dataset becomes a challenging problem. Previous work on traditional incremental maintenance can not tackle it directly. In this study, we introduce the concept of cross in order to reuse the already-known knowledge. Furthermore,we propose an efficient updating algorithm (IMCP) for maintenance of discovered spatial colocation patterns when a set of new spatial data comes.