Mining Patterns with a Balanced Interval

TL;DR

This paper introduces methods to identify patterns that occur at regular, balanced intervals in datasets, using statistical measures like standard deviation and average to improve pattern mining efficiency.

Contribution

It proposes a new approach for mining balanced interval patterns using standard deviation and average, along with a simplified pruning method for better threshold estimation.

Findings

Effective identification of balanced interval patterns.

Simplified pruning approach improves pattern selection.

Applicable to various datasets with periodic behaviors.

Abstract

In many applications it will be useful to know those patterns that occur with a balanced interval, e.g., a certain combination of phone numbers are called almost every Friday or a group of products are sold a lot on Tuesday and Thursday. In previous work we proposed a new measure of support (the number of occurrences of a pattern in a dataset), where we count the number of times a pattern occurs (nearly) in the middle between two other occurrences. If the number of non-occurrences between two occurrences of a pattern stays almost the same then we call the pattern balanced. It was noticed that some very frequent patterns obviously also occur with a balanced interval, meaning in every transaction. However more interesting patterns might occur, e.g., every three transactions. Here we discuss a solution using standard deviation and average. Furthermore we propose a simpler approach for pruning patterns with a balanced interval, making estimating the pruning threshold more intuitive.