A Graph-based Clustering Approach to Evaluate Interestingness Measures: A Tool and a Comparative Study

  title={A Graph-based Clustering Approach to Evaluate Interestingness Measures: A Tool and a Comparative Study},
  author={Hiep Xuan Huynh and Fabrice Guillet and Julien Blanchard and Pascale Kuntz and Henri Briand and R{\'e}gis Gras},
  booktitle={Quality Measures in Data Mining},
Finding interestingness measures to evaluate association rules has become an important knowledge quality issue in KDD. Many interestingness measures may be found in the literature, and many authors have discussed and compared interestingness properties in order to improve the choice of the most suitable measures for a given application. As interestingness depends both on the data structure and on the decision-maker's goals, some measures may be relevant in some context, but not in others… 

Behavior-based clustering and analysis of interestingness measures for association rule mining

By clustering based on ranking behavior, this paper highlights, and formally prove, previously unreported equivalences among interestingness measures, and shows that domain knowledge is essential to the selection of an appropriate interestingness measure for a particular task and business objective.

Characterization of Interestingness Measures Using Correlation Analysis and Association Rule Mining

This work investigates sixty-one objective interestingness measures on the pattern of A → B, to analyze their similarity and dissimilarity as well as their relationship, and selects an appropriate interestingness measure using the generated heat-map and association rules.

Finding the Most Interesting Association Rules by Aggregating Objective Interestingness Measures

This paper proposes a new approach to aggregate a set of interestingness measures using the Choquet integral as an advanced aggregation operator to extract rules satisfying many points of view.

Clustering the objective interestingness measures based on tendency of variation in statistical implications

This paper built the statistical implication data of 31 objective interestingness measures based on the examination of the partial derivatives on four parameters and built the similarity trees based on distance matrix that gave results of 31 measures clustering with two different clustering thresholds.

Classification of objective interestingness measures

This paper completes a new classification on 109 selected objective interestingness measures on 6 criterions (independence, equilibrium, symmetry, variation, description, and statistics) to obtain a general view of recent researches on the nature of the objectiveinterestingness measures.

Interestingness Measures for Association Rules in a KDD Process : PostProcessing of Rules with ARQAT Tool

This thesis proposes a new approach implemented in a new tool, ARQAT (Association Rule Quality Analysis Tool), in order to facilitate the analysis of the behavior about 40 interestingness measures, and discovered a set of correlations not very sensitive to the nature of data used, and which is called stable correlations.

Ranking Association Rules by Clustering Through Interestingness

This work proposes a process to cluster ARs based on their interestingness, according to a set of OMs, to obtain an ordered list containing the most relevant patterns, and shows that the proposed process behaves equal or better than as if the best OM had been used.

Solving the Problem of Selecting Suitable Objective Measures by Clustering Association Rules Through the Measures Themselves

This work proposes a process to solve the problem related to the identification of a suitable OM to direct the users towards the interesting patterns, and tries to reduce the exploration space to minimize the user's effort.

Ranking objective interestingness measures with sensitivity values

In this paper, we propose a new approach to evaluate the behavior of objective interestingness measures on association rules. The objective interestingness measures are ranked according to the most

Rate of change analysis for interestingness measures

The insufficiency of the existing properties in the literature to capture certain behaviors of interestingness measures is discussed and an approach where a measure is described by how it varies if there is a unit change in the frequency count is adopted.



Clustering Interestingness Measures with Positive Correaltion

A new approach implemented in a tool to select the groups or clusters of objective interestingness measures that are highly correlated in an application and relies on helping the user to selected the subset of measures that is the best adapted to discover the best rules according to his/her preferences.

Selecting the right objective measure for association analysis

Modeling of the counter-examples and association rules interestingness measures behavior

Three modelings of counter-examples are presented and the consequences of such modelizations on two important desired properties of association rules interestingness measures, that are the decrease with the number of counterexamples and the tolerance to the apparition of the first countereXamples are examined.

A Clustering of Interestingness Measures

An experimental study of the behaviour of 20 measures on 10 datasets is presented, compared to a previous analysis of formal and meaningful properties of the measures, by means of two clusterings to enhance the previous approach.

Evaluation of Interestingness Measures for Ranking Discovered Knowledge

This work theoretically and empirically evaluate thirteen diversity measures used as heuristic measures of interestingness for ranking summaries generated from databases, and describes five principles that any measure must satisfy to be considered useful forranking summaries.

Using information-theoretic measures to assess association rule interestingness

The directed information ratio (DIE) is presented, a new rule interestingness measure which is based on information theory and is a very filtering measure, which is useful for association rule post-processing.

On rule interestingness measures

What Makes Patterns Interesting in Knowledge Discovery Systems

The focus of the paper is on studying subjective measures of interestingness, which are classified into actionable and unexpected, and the relationship between them is examined.

On the robustness of association rules

A new specificity of association rule objective interestingness measures: the threshold sensitivity is proposed, allowing us to determine the number of examples that a rule can lose while remaining acceptable, for a panel of classical measures that are transformation of the confidence.

Finding Interesting Patterns Using User Expectations

In this paper, a technique to prevent the user from being overwhelmed by the large number of patterns, techniques are needed to rank them according to their interestingness, called the user-expectation method.