New Clustering and Analyzing Technique for Mining Multi-source Enriched Geo-spatial Data

Abstract

Mining multi-source enriched geo-spatial data is an important task in many application domains, such as environmental science, geographic information science, and social networks. In this paper, we propose a new density-based spatio-temporal clustering algorithm called Poly-ST-SNN and a post-processing analysis technique to extract interesting patterns and useful knowledge from multi-source enriched geo-spatial data. Poly-ST-SNN extends the well-established density-based Shared Nearest Neighbor (SNN) clustering algorithm. In contrast to previous works in this area, our approaches can cluster and analyze dynamically evolved complex enriched geo-spatial objects, i.e., polygons. We evaluate the effectiveness of our techniques through a challenging real case study involving ozone pollution events in the Houston-Galveston-Brazoria (HGB) area. The experimental results show that our approaches can discover interesting spatio-temporal patterns and useful information from multi-source enriched ozone pollution data.