Mining Interval Time Series

Data mining can be used to extensively automate the data analysis process. Techniques for mining interval time series, however, have not been considered. Such time series are common in many applications. In this paper, we investigate mining techniques for such time series. Specifically, we propose a technique to discover temporal containment relationships. An item A is said to contain an item B if an event of type B occurs during the time span of an event of type A, and this is a frequent relationship in the data set. Mining such relationships allows the user to gain insight on the temporal relationships among various items. We implement the technique and analyze trace data collected from a real database application. Experimental results indicate that the proposed mining technique can discover interesting results. We also introduce a quantization technique as a preprocessing step to generalize the method to all time series.