A new interestingness measure for maintaining specific attributes in Numerical Association Rule Mining

In Numerical Association Rule Mining, some attributes are more inclined to disappear. These attributes are inclined to disappear primarily when rules consist of a small value of support or confidence. These kinds of attributes are mainly discrete. In this paper, the problem of disappearing attributes is tackled by proposing a new interest measure that controls the specific qualities. The series of experiments underpins the rationale of the latest interest measure.