Fast mining of complex spatial co-location patterns using GLIMIT

Most algorithms for mining interesting spatial co-locations integrate the co-location / clique generation task with the interesting pattern mining task, and are usually based on the Apriori algorithm. This has two downsides. First, it makes it difficult to meaningfully include certain types of complex relationships - especially negative relationships - in the patterns. Secondly, the Apriori algorithm is slow. In this paper, we consider maximal cliques - cliques that are not contained in any other clique. We use these to extract complex maximal cliques and subsequently mine these for interesting sets of object types (including complex types). That is, we mine interesting complex relationships. We show that applying the GLIMIT itemset mining algorithm to this task leads to far superior performance than using an Apriori style approach.