SNN Neighbor and SNN Density-based co-location pattern discovery

摘要

Concerning co-location pattern mining research, the definition of co-location instance in classical algorithms is clique-based. Considering the drawbacks of this definition, this work proposes a novel definition: SNN Neighbor and SNN Density-based co-location instance. Then the paper illustrates the significance of this conception and a SNN Neighbor and SNN Density-based co-location pattern mining algorithm is realized. At last, a plenty of experimental results on synthetic and real data sets show this approach is correct and flexible, and can discovery more interesting patterns that clique-based methods fail to. Further more, our solution is faster and takes less memory consumption than traditional approaches.