Result-directed CHR Execution


The traditional execution mode of CHR is bottom-up, that is, given a goal, the result is computed by exhaustively applying rules. This paper proposes a result-directed execution mode for CHR, to be used when both the goal and the result are known, and the task is to find all corresponding derivations. Result-directed execution is needed in the context of CHRiSM, a probabilistic extension of CHR in which goals typically have a large number of possible results. The performance of result-directed execution is greatly improved by adding early-fail rules.

View Slides

Extracted Key Phrases

Cite this paper

@inproceedings{Sneyers2010ResultdirectedCE, title={Result-directed CHR Execution}, author={Jon Sneyers}, year={2010} }