Consider the Datalog program P:

T (x) e(x), T (x) a(x, y) ^ T (y) ^ b(x, z) ^ T (z)

(a) Describe, in your own words, the kind of structures that the query hT, Pi recognises. (b) Compute the semi-naive evaluation of P for the database D with the following facts: e(1) e(2) e(6) a(3, 1) a(4, 3) a(5, 3) a(7, 5) b(3, 2) b(5, 3) b(7, 6) Specify for each newly derived fact which of the rule(s) will produce it at the given point in the derivation.