![]() | ![]() | ![]() | Solutions |
p(X) <- q(X). p(Y) <- r(Y). q(a). r(b).
One model is where a, b and c all denote fred (i.e., phi(a)=fred, phi(b)=fred, and phi(c)=fred) and all relations are true (i.e., pi(p)(fred)=true, pi(p)(mary)=true, pi(q)(fred)=true, pi(q)(mary)=true, pi(r)(fred)=true, pi(r)(mary)=true).
Another model is where a denotes fred and b and c denote mary (i.e., phi(a)=fred, phi(b)=mary, and phi(c)=mary) and the relations are as follows: pi(p)(fred)=true, pi(p)(mary)=true, pi(q)(fred)=true, pi(q)(mary)=false, pi(r)(fred)=false, pi(r)(mary)=true).
Another model is where a, b and c all denote fred (i.e., phi(a)=fred, phi(b)=fred, and phi(c)=fred) and the relations are as follows: pi(p)(fred)=true, pi(p)(mary)=false, pi(q)(fred)=true, pi(q)(mary)=false, pi(r)(fred)=true, pi(r)(mary)=false).
We can select an arbitrary clause and falsify it, filling the other atoms arbitrarily. For example, any interpretation with a denoting fred (i.e., with phi(a)=fred) and pi(q)(fred)=true falsifies the third clause and isn't a model.
Similarly any interpretation with pi(p)(fred)=false and pi(p)(fred)=true falsifies the first clause, and so isn't a model, independently of the values of ]phi or the other values of pi.
So one interpretation that isn't a model is where a denotes fred and b and c denote mary (i.e., phi(a)=fred, phi(b)=mary, and phi(c)=mary) and the relations are as follows: pi(p)(fred)=false, pi(p)(mary)=false, pi(q)(fred)=false, pi(q)(mary)=true, pi(r)(fred)=true, pi(r)(mary)=false). Every clause in the knowledge base is false in this interpretation.
Any two of q(a), r(b), p(a), p(b)
Any two of r(a), q(b), r(c), q(c), p(c).
![]() | ![]() | ![]() | Solutions |