I describe the Temporal Control System (TCS), a programming system designed for building intelligent temporal monitoring programs. Empirical results from the ICU date set validate the scalable design of the TCS. I then focus on the problem of generating interval values from sample points via persistence assumptions. The TCS provides both the framework for the implementation as well as a method of calculating the "cost" of different approaches. In particular, I show that limiting the time span of a persistent interval can be very costly and then suggest how the application of symbolic abstraction can help. Further performance improvements come from the development of additional temporal abstraction techniques.
In Artificial Intelligence in Medicine: Interpreting Clinical Data. Papers from the AAAI Spring Symposium, AAAI Technical Report SS-94-01, pp. 134-138, March 1994. The full paper is available in postscript. Get Postscript. (5pp)