Measures Of Interest Research Articles

Introducing Concept of Fuzzy Support Matrix for Interestingness Measures

Introducing Concept of Fuzzy Support Matrix for Interestingness Measures

An efficient approach for discovering Graph Entity Dependencies (GEDs)

Graph entity dependencies (GEDs) are novel graph constraints, unifying keys and functional dependencies, for property graphs. They have been found useful in many real-world data quality and data management tasks, including fact checking on social media networks and entity resolution. In this paper, we study the discovery problem of GEDs—finding a minimal cover of valid GEDs in a given graph data. We formalise the problem, and propose an effective and efficient approach to overcome major bottlenecks in GED discovery. In particular, we leverage existing graph partitioning algorithms to enable fast GED-scope discovery, and employ effective pruning strategies over the prohibitively large space of candidate dependencies. Furthermore, we define an interestingness measure for GEDs based on the minimum description length principle, to score and rank the mined cover set of GEDs. Finally, we demonstrate the scalability and effectiveness of our GED discovery approach through extensive experiments on real-world benchmark graph data sets; and present the usefulness of the discovered rules in different downstream data quality management applications.

Data-Driven Insight Synthesis for Multi-Dimensional Data

Exploratory data analysis can uncover interesting data insights from data. Current methods utilize "interestingness measures" designed based on system designers' perspectives, thus inherently restricting the insights to their defined scope. These systems, consequently, may not adequately represent a broader range of user interests. Furthermore, most existing approaches that formulate "interestingness measure" are rule-based, which makes them inevitably brittle and often requires holistic re-design when new user needs are discovered. This paper presents a data-driven technique for deriving an "interestingness measure" that learns from annotated data. We further develop an innovative annotation algorithm that significantly reduces the annotation cost, and an insight synthesis algorithm based on the Markov Chain Monte Carlo method for efficient discovery of interesting insights. We consolidate these ideas into a system. Our experimental outcomes and user studies demonstrate that DAISY can effectively discover a broad range of interesting insights, thereby substantially advancing the current state-of-the-art.

Mining credible attribute rules in dynamic attributed graphs

Graph mining has numerous real-world applications. The goal is to extract interesting subgraphs or patterns in graph databases. Dynamic attributed graphs are more complex databases in which graphs change over time, and each vertex has multiple attributes. However, most algorithms only pay attention to finding rules that show relationships between nodes, and they typically have substantial limitations, such as finding rules with a single attribute or vertex and edges that do not change over time. To address these limitations, we propose an algorithm called CAR-Miner, which aims to mine credible attribute rules in dynamic attributed graphs. The algorithm incorporates a novel objective interestingness measure that is stable, anti-monotonic, and can eliminate cross-support core patterns. We conducted several experiments on real-life databases. The results show that our algorithm outperforms the state-of-the-art algorithm, especially as the number of attributes increases. Additionally, we find patterns that have practical significance and can be interpreted in various ways. Overall, CAR-Miner is a promising algorithm for mining patterns in dynamic attributed graphs, which can help researchers and practitioners identify valuable patterns that were previously difficult to discover.

A novel software defect prediction approach via weighted classification based on association rule mining

Software defect prediction technology is used to assist software practitioners in effectively allocating test resources and identifying hidden defects in a timely manner. However, the prediction of defect-prone software using association rule mining algorithms is limited because of the unbalanced distribution of defect data. Furthermore, although the existing weighted association rule mining approach considers item strength, the weight calculation still relies on expert experience and lacks fine granularity. We propose a novel software defect prediction approach based on mutual information and correlation coefficient weighted class association rule mining (MCWCAR). The MCWCAR model employs a cost-sensitive strategy and generates frequent itemsets according to three mining objectives while maintaining the original item distribution: defective class rules, non-defective class rules, and feature association relationships. During the weighted frequent itemset mining process, it combines feature selection and itemset screening to determine the appropriate feature combination through mutual information weighted support. Meanwhile, the correlation coefficient is applied to accurately depict the correlation between feature items and defect classes, serving as the weight to mine class association rules. Additionally, to ensure that interestingness measures have asymmetry and effectively represent negative associations under the condition of class imbalance, the addedvalue is adopted in the filtering association rules. We conducted experiments on 27 open-source datasets and evaluated the performance differences between MCWCAR and state-of-the-art baseline classifiers. Experimental results demonstrate that the proposed algorithm significantly outperforms other baselines in terms of Balance, Gmean, MCC, and F-measure.

An interestingness measure for knowledge bases

Association rule mining and logical rule mining both aim to discover interesting relationships in data or knowledge. In association rule mining, relationships are identified based on the occurrence of items in a dataset, while in logical rule mining, relationships are determined based on logical relationships between atoms in a knowledge base. Association rule mining has been widely studied in transactional databases, mainly for market basket analysis. Confidence has become the most widely used interesting measure to assess the strength of a rule. Many other interestingness measures have been proposed since confidence can be insufficient to filter negatively associated relationships. Recently, logical rule mining has become an important area of research, as new facts can be inferred by applying discovered logical rules. They can be used for reasoning, identifying potential errors in knowledge bases, and to better understand data. However, there are currently only a few measures for logical rule mining. Furthermore, current measures do not consider relations that can have several objects, called quasi-functions, which can dramatically alter the interestingness of the rule. In this paper, we focus on effectively assessing the strength of logical rules. We propose a new interestingness measure that takes into account two categories of relations, functions and quasi-functions, to assess the degree of certainty of logical rules. We compare our proposed measure with a widely used measure on both synthetic test data and real knowledge bases. We show that it is more effective in indicating rule quality, making it an appropriate interestingness measure for logical rule evaluation.

Boosting the Learning for Ranking Patterns

Pattern mining is a valuable tool for exploratory data analysis, but identifying relevant patterns for a specific user is challenging. Various interestingness measures have been developed to evaluate patterns, but they may not efficiently estimate user-specific functions. Learning user-specific functions by ranking patterns has been proposed, but this requires significant time and training samples. In this paper, we present a solution that formulates the problem of learning pattern ranking functions as a multi-criteria decision-making problem. Our approach uses an analytic hierarchy process (AHP) to elicit weights for different interestingness measures based on user preference. We also propose an active learning mode with a sensitivity-based heuristic to minimize user ranking queries while still providing high-quality results. Experiments show that our approach significantly reduces running time and returns precise pattern ranking while being robust to user mistakes, compared to state-of-the-art approaches.

Bounding box representation of co-location instances for Chebyshev and Manhattan metrics

Co-location Pattern Mining (CPM) is the task of discovering sets of spatial features (object types) whose instances are frequently located close to each other in space. Popular co-location discovery methods consist of iteratively: (1) generating co-location candidates, (2) determining instances of these candidates and calculating a measure of potential interestingness, and (3) determining the set of co-locations based on that measure. In this paper, we focus on the second step, as it is the most time-consuming element of CPM. We assume that the distance function is either the Chebyshev or the Manhattan metric. We provide an instance identification method that is characterized by a lower complexity than the state-of-the-art approach. In particular, (1) we introduce a new representation of co-location instances based on bounding boxes, (2) we formulate and prove several theorems regarding such a representation that can improve instances identification step, (3) we provide a novel algorithm that uses the above-mentioned theorems, and (4) we analyze its complexity. To verify our approach, we performed a series of experiments using two real-world datasets.

Interestingness Measure of Graph-Pattern Association Rules for Knowledge Bases

Interestingness Measure of Graph-Pattern Association Rules for Knowledge Bases

Scenario-based analysis for discovering relations among interestingness measures

Many interestingness measures have been proposed for mining meaningful association rules among two events in the form of A→B, but their characteristics and semantic similarity relations have not been comprehensively investigated. This paper presents a scenario-based approach for characterizing sixty-one commonly used measures and revealing their relationships in three steps. The first step generates a set of 969 three-probability scenarios, S={s|s=(p(A),p(B),p(A,B))∧p(A),p(B),p(A|B)∈[0,1]∧p(A,B)⩽min(p(A),p(B))}, in consideration of all possible situations in the range of 0.0 to 1.0 with a step of 0.05, excluding infinity and not-a-number cases. In the second step, 937,992 pairs of scenarios are enumerated, and for each scenario pair s1 and s2, the values of a measure (M) of s1 and s2, i.e., M(s1) and M(s2), are compared with the result of greater-than (M(s1)>M(s2)), smaller-than (M(s1)<M(s2)), or equal-to (M(s1)=M(s2)) for characterizing the measure. The final step is based on three types of relations: (1) behavior-based, (2) correlation-based, and (3) association-based similarity relations. The behavior of measures is depicted using nine common algebraic/statistical properties and four special condition properties, i.e., zero, min–max, infinity, and not-a-number of the measures. Similarities among the measures can be examined by grouping measures based on their properties. With three correlation functions, i.e., correlation coefficient, joint entropy, and mutual information, a correlation analysis was performed to discover relations among interestingness measures in the form of dendrograms and clusters with thresholding. Finally, the details of the relations among these interestingness measures are explored with association rule mining. Besides support, confidence, and lift, we propose five types of rules, i.e. same-directionrule (S-rule), opposite-direction rule (O-rule), equal-both rule (E-rule), equal-left rule (EL-rule),and equal-right rule (ER-rule) for a five-gradient comparison of any two measures to outline their similarities and dissimilarities in five directions.

Impact of COVID-19 on Personal Life: an Analysis Based on the Data Centric Approach

COVID-19 is the most used and significant word in this year. It has taken many lives, created fear among people and changed people’s and country’s perspective and decision. The virus has kept people isolated in their homes, and borders have been blocked. The third wave of Covid-19 variants has finally come to an end. Based on an online questionnaire, this paper investigated the impacts of Covid-19 on people’s lives for four essential daily necessities: education, social, economic, and psychological. As the pandemic interrupted the normal world’s circumstances, we tried to determine the association between various factors related to Bangladeshi people in this study. This research is essential because the findings help stakeholders and decision makers decide what to do in the sectors of education, the economy, social, and psychology. This study introduces a new interestingness measure, the “IM-Score,” which aims to discover the association between various parameters. Covid-19 impedes 25.6 percent of educational factors, 23.5 percent of social concerns, 25.9 percent of economic life, and 25.0 percent of psychological factors among the four most basic demands of people’s daily lives, according to the findings. Finally, we concluded that people are concerned about the period of online classes, tuition expenses, financial situation, social interaction, and mental health.

An Improved Evaluation Methodology for Mining Association Rules

At present, association rules have been widely used in prediction, personalized recommendation, risk analysis and other fields. However, it has been pointed out that the traditional framework to evaluate association rules, based on Support and Confidence as measures of importance and accuracy, has several drawbacks. Some papers presented several new evaluation methods; the most typical methods are Lift, Improvement, Validity, Conviction, Chi-square analysis, etc. Here, this paper first analyzes the advantages and disadvantages of common measurement indicators of association rules and then puts forward four new measure indicators (i.e., Bi-support, Bi-lift, Bi-improvement, and Bi-confidence) based on the analysis. At last, this paper proposes a novel Bi-directional interestingness measure framework to improve the traditional one. In conclusion, the bi-directional interestingness measure framework (Bi-support and Bi-confidence framework) is superior to the traditional ones in the aspects of the objective criterion, comprehensive definition, and practical application.

A Statistical Generation Method of Association Rules by Inference of Diagnostic Test Index in Data Mining

In this paper, we investigated the possibility of using Youden s index, which is a diagnostic test index used in diagnostic medicine, as an interestingness measure among objective thresholds in association rules. First, we looked at the relationship of confidence measures for the sensitivity and specificity that constitute the Youden index. We also examined the relationship between Youden index and confidence measures by expressing the Youden index formula using various types of confidence, and presented the confidence interval of the Youden index. In addition, after grasping the meaningful properties as interestingness measure, we proved whether the condition of interestingness measure proposed by Piatetsky-Shapiro (1991) was satisfied. By simulation study, if the lower limit of the Youden index has a positive value, it is concluded that there is a positive association rule. If the upper limit of the Youden index has a negative value, it is determined that there is a negative association. If the lower limit of the Youden index has a negative value but the upper limit is positive, then it would be desirable to reserve whether or not there is a association rule.

Conceptual Coverage Driven by Essential Concepts: A Formal Concept Analysis Approach

Formal concept analysis (FCA) is a mathematical theory that is typically used as a knowledge representation method. The approach starts with an input binary relation specifying a set of objects and attributes, finds the natural groupings (formal concepts) described in the data, and then organizes the concepts in a partial order structure or concept (Galois) lattice. Unfortunately, the total number of concepts in this structure tends to grow exponentially as the size of the data increases. Therefore, there are numerous approaches for selecting a subset of concepts to provide full or partial coverage. In this paper, we rely on the battery of mathematical models offered by FCA to introduce a new greedy algorithm, called Concise, to compute minimal and meaningful subsets of concepts. Thanks to its theoretical properties, the Concise algorithm is shown to avoid the sluggishness of its competitors while offering the ability to mine both partial and full conceptual coverage of formal contexts. Furthermore, experiments on massive datasets also underscore the preservation of the quality of the mined formal concepts through interestingness measures agreed upon by the community.

A novel imbalanced data classification approach for suicidal ideation detection on social media

Suicide has become a serious social health issue in modern society. Suicidal ideation is people’s thoughts about committing or planning suicide. Many factors, such as long-term exposure to negative feelings or life events, can lead to suicidal ideation and suicide attempts. Among these approaches to suicide prevention, early detection of suicidal ideation is one of the most effective ways. Using social networking services provides a platform for people to express their sufferings and feelings in the real world, which provides a source for a deeper investigation into models and approaches for the detection of suicidal intent to enable prevention. This paper addresses the early detection of suicide ideation through the associative classification approach applied to Twitter social media. However, since the number of suicide intention tweets is tiny compared to the number of all the tweets, this leads us to an imbalanced classification problem, in which, the minority class (suicide intention) is more important than the majority class (no suicide intention). In such a situation, classical classifiers usually yield very inaccurate results regarding minor classes, since they can easily discover rules predicting the majority class and overlook those related to the minor. This paper aims to contribute to this line of research by introducing a new interestingness measure to enhance the classification process. This measure highlights the two classes regardless of their imbalanced distribution. Carried out experiments proved that the adapted CBA outweighs in terms of prediction accuracy the original one, and other pioneering baseline classification approaches.

Mining predicate rules without minimum support threshold

Association rule mining (ARM) is used for discovering frequent itemsets for interesting relationships of associative and correlative behaviors within the data. This gives new insights of great value, both commercial and academic. The traditional ARM techniques discover interesting association rules based on a predefined minimum support threshold. However, there is no known standard of an exact definition of minimum support and providing an inappropriate minimum support value may result in missing important rules. In addition, most of the rules discovered by these traditional ARM techniques refer to already known knowledge. To address these limitations of the minimum support threshold in ARM techniques, this study proposes an algorithm to mine interesting association rules without minimum support using predicate logic and a property of a proposed interestingness measure (g measure). The algorithm scans the database and uses g measure’s property to search for interesting combinations. The selected combinations are mapped to pseudo-implications and inference rules of logic are used on the pseudo-implications to produce and validate the predicate rules. Experimental results of the proposed technique show better performance against state-of-the-art classification techniques, and reliable predicate rules are discovered based on the reliability differences of the presence and absence of the rule’s consequence.

Spark solutions for discovering fuzzy association rules in Big Data

The high computational impact when mining fuzzy association rules grows significantly when managing very large data sets, triggering in many cases a memory overflow error and leading to the experiment failure without its conclusion. It is in these cases when the application of Big Data techniques can help to achieve the experiment completion. Therefore, in this paper several Spark algorithms are proposed to handle with massive fuzzy data and discover interesting association rules. For that, we based on a decomposition of interestingness measures in terms of α-cuts, and we experimentally demonstrate that it is sufficient to consider only 10 equidistributed α-cuts in order to mine all significant fuzzy association rules. Additionally, all the proposals are compared and analysed in terms of efficiency and speed up, in several datasets, including a real dataset comprised of sensor measurements from an office building.

MoMAC: Multi-objective optimization to combine multiple association rules into an interpretable classification

A crucial characteristic of machine learning models in various domains (such as medical diagnosis, financial analysis, or real-time process monitoring) is the interpretability. The interpretation supports humans in understanding the meaning behind every single prediction made by the machine, and enables the user to assess trustworthiness before acting on the predictions. This article presents our work in building an interpretable classification model based on association rule mining and multi-objective optimization. The classification model itself is a rule list, making a single prediction based on multiple rules. The rule list consists of If ... THEN statements that are understandable to humans. We choose these rules from a large set of pre-mined rules according to an interestingness measure which is formulated as a function of basic probabilities related to the rules. We learned the interestingness measure through multi-objective optimization, concentrating on two objectives: the classifier’s size in terms of number of rules and prediction accuracy. The model is called MoMAC, “Multi-Objective optimization to combine Multiple Association rules into an interpretable Classification”. The experimental results on benchmark datasets demonstrate that MoMAC outperforms other existing rule-based classification methods in terms of classification accuracy.

Optimization of association rules using hybrid data mining technique

Software quality has been the important area of interest for decades now in the IT sector and software firms. Defect prediction gives the tester the pointers as to where the bugs will most likely be hidden in the software product. Identifying and reporting the defect probe areas is the main job of software defect prediction techniques. Early detection of software defects during Software Development Life Cycle could lead to a reduction in cost of development, time involved in further testing activities and rework effort post-production and maintenance phase, thus resulting in more reliable software. Software metrics can be used for developing the defect prediction models. Several data mining techniques can be applied on the available open-source software datasets. These datasets are extracted from software programs. Such datasets made publicly available by National Aeronautics and Space Administration for their various softwares have been extensively used in software engineering-related research activities. These datasets contain information on associated Software Metrics at module level. The proposed idea is a novel hybrid data mining technique consisting of Clustering and Modified Apriori Algorithm that results in improved efficiency and reliability of Software Defect Prediction. This technique works by reducing the number of association rules generated. The results are achieved by using interestingness measure called spread. The paper also does a comparative analysis of the results obtained from the novel technique with the existing hybrid technique of Clustering and Apriori.

A novel interestingness measure based on fusion model for association rules mining

Aiming at the shortcomings of the traditional "support-confidence" association rules mining framework and the problems of mining negative association rules, the concept of interestingness measure is introduced. Analyzed the advantages and disadvantages of some commonly used interestingness measures at present, and combined the cosine measure on the basis of the interestingness measure model based on the difference idea, and proposed a new interestingness measure model. The interestingness measure can effectively express the relationship between the antecedent and the subsequent part of the rule. According to this model, an association rules mining algorithm based on the interestingness measure fusion model is proposed to improve the accuracy of mining. Experiments show that the algorithm has better performance and can effectively help mining positive and negative association rules.