Rule lift

The lift of a rule is the strength divided by the RHS coverage proportion. This indicates how much more frequent is the RHS if the LHS occurs than normal.

For example, suppose that there are 1000 examples, the LHS covers 200 examples, the RHS covers 100 examples, and the RHS covers 50 of the examples covered by the LHS. The strength is 50/200 = 0.25. The RHS cover proportion is 100/1000 = 0.1. The lift is 0.25/0.1 = 2.5.

Note, when m-estimates are used, a lift estimate is displayed in place of the lift of the rule on the data.