Association rules over interval data

Abstract

We consider the problem of mining association rules over interval data (that is, ordered data for which the separation between data points has meaning). We show that the measures of what rules are most important (also called rule interest ) that are used for mining nominal and ordinal data do not capture the semantics of interval data. In the presence of interval data, support and confidence are no longer intuitive measures of the interest of a rule. We propose a new definition of interest for association rules that takes into account the semantics of interval data. We developed an algorithm for mining association rules under the new definition and overview our experience using the algorithm on large real-life datasets.

Characterisation of gradual itemsets based on mathematical morphology tools

“…Quantitative association rules Quantitative association rules are an extension of classical association rules to numerical attributes [10,11]: in this case indeed, an item cannot be defined as an attribute value, because the notion of occurrence frequency for a numerical value is not relevant. An item is defined as a couple made of an attribute with an interval, e.g.…”

“…defined as equi-width or equidepth intervals [10,12,11]. Intervals of interest are then identified with the Apriori algorithm, applied to extended data where binary features for each interval are added to indicate whether the numerical value of an attribute belongs to the corresponding interval.…”

Harmonized Scheme for Data Mining Technique to Progress Decision Support System in an Uncertain Situation

“…[6] Define sequence discovery as "a sequential technique is a given set of sequences find the complete set of frequent subsequences". Clustering is "the process of organizing objects into groups whose members are alike in some way" [7]. So, it deals with finding the internal structure in a collection of data, as shown in figure 3.…”

“…Simple graphical for clustering data [7] [8] Define that "Clustering involves identifying a finite set of categories or segments "clusters" to describe the data according to a certain metric". [9] Define that "Clustering enables to find specific discriminative factors or attributes for the studied data.…”

Implication-Based Fuzzy Association Rules

“…Thus, not only the rating but also the mining of quantitative rules can be reduced to the mining of binary association rules, by simply transforming the numerical data into binary data [18,22]. Still, finding a useful transformation (binarization) of the data is a non-trivial problem by itself which affects both, the efficiency of subsequently applied mining algorithms and the potential quality of discovered rules.…”