Query Optimization: Ordering Rules?

We discuss a new heuristic optimization concept for query evaluation. It is based on the idea of assigning an "optimization potential" as a special weighting factor expressing the quality to all transformation rules present in our term rewriting system. In this way, the idea is a refinement of existing proposals talking about "promise functions", "expected cost factors", and "rule selection". But as will be shown, our ordering of rules is justified in much more detail. The rule ordering is used to decide which rule has to be tried next during rewriting. Thus, minimizing the optimization effort of term rewriting is achieved by minimizing the matching effort. This is an optimization of an optimization, consequently in the true sense of the word we would have to talk about metaoptimization. We omit this for our own convenience. Keywords: Query evaluation, query optimization, optimization concept, heuristic optimization, rule--based rewriting, rule ordering, adaptation, classi..