Optimizing online spatial data analysis with sequential query patterns

The exponential growth of the usage of geographic information services leads to an extensive popularity on spatial data analysis in both industrial and academic communities. However, it is quite challenging for users to efficiently analyze and quickly understand the spatial data due to the inherently complex and dynamic nature of GIS applications. To address the challenges, this paper presents an approach to optimize the online spatial data analysis by mining the sequential query patterns from the user query logs of GIS applications. The sequential query patterns are used to automatically generate the query template, from which the users are able to quickly compose new queries. The sequential query patterns contribute to the workflow construction for complex spatial data analysis tasks as well. Our proposed approach takes advantage of the generated workflow to parallelize the independent spatial analysis tasks. As a result, the throughput of our system has been increased greatly and more efficient geographic information services are made available to the users. We present a case study to demonstrate the efficiency and effectiveness of the proposed approach by integrating two software systems at Florida International University (FIU): TerraFly (an GIS application) and FIU-Miner (a distributed data mining framework).