A Dynamic Thresholding Technique In Spatially Co-Located Objects Mining From Vehicle Moving Data

Abstract

Co-location pattern discovery is intended towards the processing data with spatial contexts to discover classes of spatial objects that are frequently located together. The existing moving vehicle location prediction technique not analyses the moving vehicles co-location instance. So, we improve the previous technique process by mining spatially co-located moving objects using spatial data mining techniques. Initially, the neighbour relationship is computed by the prim's algorithm. After that, the candidate co-locations are pruned according to the presence of candidate co-location in the input data and the final stage of co-location instances selection is performed by compute neighbourhood and node membership functions. The values obtained using neighbourhood membership function is compared with the dynamic threshold values. The co-location instances are selected which satisfy the dynamic threshold value. Moreover, the proposed co-location pattern mining with dynamic thresholding technique is compared with the existing co-location pattern mining technique.