Mining data quality rules based on T-dependence
Toon Boeckling, Antoon Bronselaer, Guy De Tré · 2019
Since their introduction in 1976, edit rules have been a standard tool in statistical analysis.Basically, edit rules are a compact representation of non-permitted combinations of values in a dataset.In previous work, edit rules are mined automatically as valuecombinations showing strong negative correlation tested under stochastic independence using the traditional notion of lift.In this paper, we generalize the traditional notion of lift to that of T -lift, where stochastic independence is generalized to T -dependence.We show several interesting properties of edit rules under different T -lift measures and proof that edit rules under the minimum tnorm can be computed efficiently by use of frequent pattern trees.Experiments support this result and show that there is a weak to medium correlation in the rank order of edit rules obtained under the minimum and product t-norms.