The Implementation of Spatial Co-location Mining and Its Application for POIs in Beijing

With the popularity and widening application of GPS vehicle navigation and Location-based service (LBS), a large amount of POI (points of interest) data has become available. The application of spatial data mining with the booming spatial data can reveal implied and interesting patterns. In this paper, we have designed the data structure for the basic definitions proposed in the co-location mining task and implemented the co-location mining algorithm with Java. Then we apply the co-location mining method in a real life case with POI in city center of Beijing. From the study case, we found some useful and interesting co-location rules among catering, commerce, and entertainment industry. These implied rules can be meaningful to city planning, commercial activities and analysis of people's consumption behavior.