A Design for Knowledge Discovery Decision Support Systems (KD-DSS) With Spatio-Temporal Data

Examining relationships in spatio-temporal data has lead to the construction of a decision support system specializing in data-mining. Several steps are taken with the system architecture to ensure a robust, easy to use online interface in a distributed environment, including the implementation of developing technologies and effective knowledge discovery algorithms. Drought risk management is the tested application of the design, looking for significant rules between global events, like La Niña, and local weather data. The National Agricultural Decision Support System embodies the traits required by drought risk management, with its application-independence and functional data-mining support. Furthermore, its architecture layers handle the complications arising from the spatio-temporal data. The algorithms REAR and MOWCATL provide knowledge discovery and are shown to provide interesting results for isolated weather stations. The decision support system is an important step towards proactive drought management.