Fast Generation of Best Interval Patterns for Nonmonotonic Constraints
This work addresses pattern mining for researchers dealing with nonmonotonic constraints, offering an incremental improvement in efficiency for specific domain applications.
The paper tackles the challenge of exponential pattern explosion in pattern mining by introducing projection-antimonotonicity and the $θ$-$Σøφια$ algorithm to efficiently generate best patterns for nonmonotonic constraints like stability and $Δ$-measure, applied to interval tuple datasets, showing experimental advantages over postfiltering approaches.
In pattern mining, the main challenge is the exponential explosion of the set of patterns. Typically, to solve this problem, a constraint for pattern selection is introduced. One of the first constraints proposed in pattern mining is support (frequency) of a pattern in a dataset. Frequency is an anti-monotonic function, i.e., given an infrequent pattern, all its superpatterns are not frequent. However, many other constraints for pattern selection are not (anti-)monotonic, which makes it difficult to generate patterns satisfying these constraints. In this paper we introduce the notion of projection-antimonotonicity and $θ$-$Σøφια$ algorithm that allows efficient generation of the best patterns for some nonmonotonic constraints. In this paper we consider stability and $Δ$-measure, which are nonmonotonic constraints, and apply them to interval tuple datasets. In the experiments, we compute best interval tuple patterns w.r.t. these measures and show the advantage of our approach over postfiltering approaches. KEYWORDS: Pattern mining, nonmonotonic constraints, interval tuple data