Comparing Rete and sequential algorithmes

Question

Which (Rete or sequential) algorithm is more efficient when rules data is known upfront and not changing?

Answer 1

If your data is not modified and there is no chaining or inference then Rete is not necessary - Rete is better for systems where data changes incrementally over time.

There are other algorithms like Phreak that support inference and sequential modes, that offers a reasonable compromise between the two. http://blog.athico.com/2013/11/rip-rete-time-to-get-phreaky.html

The biggest difference would probably come from code generated rules vs a general engine processing a data structure to represent rules. It's easier to code generate for sequential execution. Although again you could probably code generate for the Phreak algorithm's sequential mode - which would close the gap.