Pattern mining for query answering in marine sensor data

Abstract

An integrated pattern mining technique for query answering is proposed for marine sensor data. In pattern query, we adopt the dynamic time warping (DTW) method and propose the use of a query relaxation approach in finding similar patterns. We further calculate prediction from discovered similar patterns in marine sensor data. The predictive values are then compared with the forecast from hydrodynamic model data. In addition, we present query answering using a clustering technique. Finally, we show implementation results in a marine sensor network deployed in the South East of Tasmania, Australia.