Incremental and parallel spatial association mining

Spatial association mining has been used for discovering frequent spatial association patterns from large static spatial databases. When a large spatial database is updated, it is computationally expensive to redo the pattern discovery process for the updated database. This work presents the problem of finding spatial association patterns incrementally from evolving databases which are constantly updated with fresh data. The proposed method is implemented on the MapReduce framework for large-scale spatial data processing, and empirically evaluated. The developed algorithm shows substantial performance improvements when compared with an iterative and non-incremental spatial association mining algorithm.