This paper discusses retrieval of a massive amount of sensor data using an indexing mechanism called I-Tree. Based on a dimensionality reduction technique for time-series data, we introduce a symbolic extensible representation of spatial time series data, and use a relevant tree structure to manage spatial time series data efficiently. I-Tree facilitates analysis and detection of environmental phenomena using heterogeneous sensor data from multiple locations.
|Translated title of the contribution||I-Tree: Supporting Spatial Time-Series Queries for Heterogeneous Sensor Data|
|Number of pages||6|
|Journal||IEICE technical report|
|Publication status||Published - Jun 21 2010|