FOIL's Information Gain

Name	FOIL's Information Gain
Description	The measure, similar to information gain, used by FOIL is Gain(R0, R1) := t * ( log2(p1/(p1+n1)) - log2(p0/(p0+n0)) ). R0 denotes a rule before adding a new literal. R1 is an extesion of R0. Tupels are bindings of variables, restricted to constants occuring in the read examples and background knowledge. A tupel is covered by a rule in this case, if it fulfills all of its preconditions. p0 denotes the number of positive tupels, covered by R0, p1 the number of positive tupels, covered by R1. n0 and n1 are the number of negative tupels, covered by the according rule. t is the number of positive tupels, covered by R0 as well as by R1. If a literal is added, that contains a new variable, then a positive tupel covered by R0 is regarded to be covered by R1 as well, if there exists an extension of the tupel, just adding a constant as a new variable binding, so the extended tupel is covered by R1.
Algorithm	FOIL