Mining Hidden Association Rules from Real-Life Data

Mining Hidden Association Rules from Real-Life Data

Marco-Antonio Balderas Cepeda (Universidad de Granada, Spain)
DOI: 10.4018/978-1-60566-754-6.ch011
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Association rule mining has been a highly active research field over the past decade. Extraction of frequency-related patterns has been applied to several domains. However, the way association rules are defined has limited people’s ability to obtain all the patterns of interest. In this chapter, the authors present an alternative approach that allows us to obtain new kinds of association rules that represent deviations from common behaviors. These new rules are called anomalous rules. To obtain such rules requires that we extract all the most frequent patterns together with certain extension patterns that may occur very infrequently. An approach that relies on anomalous rules has possible application in the areas of counterterrorism, fraud detection, pharmaceutical data analysis and network intrusion detection. They provide an adaption of measures of interest to our anomalous rule sets, and we propose an algorithm that can extract anomalous rules as well. Their experiments with benchmark and real-life datasets suggest that the set of anomalous rules is smaller than the set of association rules. Their work also provides evidence that our proposed approach can discover hidden patterns with good reliability.
Chapter Preview
Top

Background

Association rules (Agrawal et al., 1993) is a term that refers to patterns that relate sets of items. For a given set of k items, known as a k-itemset, an association rule has two components: an antecedent and a consequent, both of which may themselves involve k-itemsets. The association rule can be evaluated by two measures; support and confidence. Support quantifies the statistical significance of the pattern, and it is defined as the probability of a certain k-itemset being part of the dataset. The confidence, defined as the conditional probability between the pattern and the antecedent of the association rule, provides evidence of the relevance of a certain rule. An association rule is confirmed if its support and confidence values are greater than or equal to certain thresholds specified by the user.

Unlike the traditional frequent itemset framework that is used to obtain association rules, research into the detection of rare and hidden patterns must consider the specific definitions that pertain to each case. Several recent publications have explored the modeling of patterns that reflect deviations from a common behavior.

Complete Chapter List

Search this Book:
Reset