Adaptive query processing for time-series data

Traditional query process for time-series data transforms data from the time domain into the frequency domain. It is less controllable by the end users, therefore isn't well suited for queries that find patterns with many dynamically specified user constraints. For these queries we present a method to search timeseries data which first transforms time sequences into symbol strings using change ratio between contiguous data points in time series. Next, a suffix tree is built to index all suffixes of the symbol strings. The focus of this paper is to demonstrate how this method can adapt to the processing of the dynamically constrained time-series queries.