Mining Spatial and Spatio-temporal Patterns in Scientific Data

This paper focusses on designing and applying data mining techniques to analyze spatial and spatiotemporal data originated in scientific domains. Data mining is the process of discovering hidden and meaningful knowledge in a data set. It has been successfully applied to many real-life problems, for instance, web personalization, network intrusion detection, and customized Marketing. This paper mainly works on developing data mining techniques to analyze spatial and spatiotemporal data produced in different scientific domains where spatial data is data pertaining to the location, shape, and relationships on features. Mining spatial and spatio-temporal relationships in scientific data is very challenging because of issues like shape and size of the datasets and problems in incorporating temporal information. To address these challenges, a generalized framework is proposed. This framework consists on robust techniques for modeling shape and extent of features. Then fast algorithms are developed for extracting frequent spatial object interactions through the design of appropriate distance functions and interaction types. After an effective approach for mining spatio-temporal episodes of spatial object association patterns are developed. Finally all these approaches are empirically evaluated on real case study applications and show that the algorithms scale well and are capable of processing large datasets. Detecting Spatio-temporal Outliers in Climate Dataset: A Method Study