Hey lads,
I want to make sure I got something correctly
In page 215, in the Simple Conversion Procedure to CNF, we start with the statement
[B1,1 iff (P1,2 OR P2,1)]
, then we end up with
[(!B1,1 OR P1,2 OR P2,1) AND (!P1,2 OR B1,1) AND (!P2,1 OR B1,1)] which is said 'can be used as an input to a resolution procedure...
My problem is, what do we get if we DO input it to a resolution procedure?
- taking the first two: (!B1,1 OR P1,2 OR P2,1) AND (!P1,2 OR B1,1)
- we can infer (P1,2 OR P2,1 OR !P1,2) by removing !B1,1 with
B1,1
- then by simple logic, we can remove P1,2 with !P1,2 to get
(P2,1)
- now by combining the result of the first two (P2,1) with the next clause (!P2,1 OR B1,1)
- we can infer B1,1, which is already what we know
So, as far as I can understand, that is what is meant by
Refutation Completeness property of
Resolution...
Sa7?
--
Cheers,
Seth
<A. Helmy>
==================
One Life... LIVE It
==================