Consecutive Support: Better Be Close!

TL;DR

This paper introduces a new measure called consecutive support that emphasizes patterns occurring frequently and closely in transaction streams, enabling faster pattern discovery and revealing interesting phenomena across various fields.

Contribution

It proposes the concept of consecutive support and integrates it into the Eclat algorithm, enhancing pattern mining by considering temporal proximity and frequency.

Findings

Synthetic examples demonstrate the effectiveness of the approach.

The method can uncover meaningful patterns in bioinformatics and other domains.

Faster pattern discovery through search space pruning.

Abstract

We propose a new measure of support (the number of occur- rences of a pattern), in which instances are more important if they occur with a certain frequency and close after each other in the stream of trans- actions. We will explain this new consecutive support and discuss how patterns can be found faster by pruning the search space, for instance using so-called parent support recalculation. Both consecutiveness and the notion of hypercliques are incorporated into the Eclat algorithm. Synthetic examples show how interesting phenomena can now be discov- ered in the datasets. The new measure can be applied in many areas, ranging from bio-informatics to trade, supermarkets, and even law en- forcement. E.g., in bio-informatics it is important to find patterns con- tained in many individuals, where patterns close together in one chro- mosome are more significant.