Generate Association Rules Research Articles

Reduction of Redundant Rules in Association Rule Mining-Based Bug Assignment

Bug triaging is a process to decide what to do with newly coming bug reports. In this paper, we have mined association rules for the prediction of bug assignee of a newly reported bug using different bug attributes, namely, severity, priority, component and operating system. To deal with the problem of large data sets, we have taken subsets of data set by dividing the large data set using [Formula: see text]-means clustering algorithm. We have used an Apriori algorithm in MATLAB to generate association rules. We have extracted the association rules for top 5 assignees in each cluster. The proposed method has been empirically validated on 14,696 bug reports of Mozilla open source software project, namely, Seamonkey, Firefox and Bugzilla. In our approach, we observe that taking on these attributes (severity, priority, component and operating system) as antecedents, essential rules are more than redundant rules, whereas in [M. Sharma and V. B. Singh, Clustering-based association rule mining for bug assignee prediction, Int. J. Business Intell. Data Mining 11(2) (2017) 130–150.] essential rules are less than redundant rules in every cluster. The proposed method provides an improvement over the existing techniques for bug assignment problem.

연관성 규칙 마이닝에서의 수정된 대칭적 J 측도의 활용 방안

오늘날 정보통신기술의 발달과 모바일 기기 이용의 확산에 의해 데이터가 폭증하게 됨으로써 빅 데이터의 환경이 도래하였으며, 이를 분석하기 위해 여러 가지 데이터 마이닝 기법이 개발되고 있다(Park, 2017a). 데이터 마이닝은 빅 데이터에 내재되어 있는 규칙이나 정보를 탐색하는 기술로, 연관성 규칙, 의사결정나무, 군집분석, 신경망 등 여러 가지 기법이 있다. 본 논문에서는 연관성 규칙 마이닝에서 이용되는 객관적 흥미도 측도 중에서 교차 엔트로피로 불리는 J 측도를 응용한 측도를 제안하고자 한다. 연관성 규칙 생성 시 신뢰도의 크기에 의해 규칙 생성 여부를 판단하게 되는데 신뢰도는 항목의 위치가 바뀌면 신뢰도의 크기가 달라지므로 두 값의 크기가 많은 차이가 나면 어느 것을 기준으로 연관성 규칙 생성 여부를 결정해야 할지 판단하기가 곤란하다(Park, 2016). 이러한 문제를 해결하기 위해 본 논문에서는 수정된 대칭적 J 측도를 제안하고자 한다. 또한 실제 사례와 모의실험 자료를 이용하여 J 측도와 관련된 기존의 측도들과의 비교를 통해 수정된 대칭적 J 측도의 유용성에 대해 탐색하고자 한다.Today, with the development of information and communication technology and the spread of mobile devices, data has been exploding and big data environment has arrived (Park, 2017a). Various data mining techniques have been developed to analyze big data. Data mining is a technique to search for rules or information inherent in big data. There are various techniques such as neural networks, association rules, decision trees, and cluster analysis. In this paper, we propose a measure applying a J measure called cross entropy among the objective interest measures used in association rule mining. When the association rule is created, it is judged whether or not the association rule is generated by the size of confidence. When the position of the item is changed, the confidence degree is changed. Therefore, if the difference of two confidence values is large, it is difficult to judge which one to decide whether to generate association rules based on which one. To solve this problem, we propose a modified symmetric J measure. We also use examples to explore the usefulness of this measure through comparison with existing measures related to J measures.

An Approach to Improve Generation of Association Rules in Order to Be Used in Recommenders

This paper introduces a method for improving the association rules that will also help improve the performance of recommender systems. Combining sampling and parallelism in the process, the proposed method, in addition to help perform the process more quickly, better quality rules will be generated. The proposed method and the dataset will be segmented into a number of smaller sections based on clustering items, and each section is separately sampled. Frequent itemsets and association rules in any section of the selected samples will be found and by gathering and analyzing the results, the quality and number of rules will be evaluated. One of the innovations of this article is the method of sampling of the database. Instead of random sampling, information about the best users and items was isolated. After determining the precise amount of support, the authors extracted the frequent and favorable rules from the selected sample.

Exploring associations of clinical and social parameters with violent behaviors among psychiatric patients

Evidence has revealed interesting associations of clinical and social parameters with violent behaviors of patients with psychiatric disorders. Men are more violent preceding and during hospitalization, whereas women are more violent than men throughout the 3days following a hospital admission. It has also been proven that mental disorders may be a consistent risk factor for the occurrence of violence. In order to better understand violent behaviors of patients with psychiatric disorders, it is important to investigate both the clinical symptoms and psychosocial factors that accompany violence in these patients. In this study, we utilized a dataset released by the Partners Healthcare and Neuropsychiatric Genome-scale and RDoC Individualized Domains project of Harvard Medical School to develop a unique text mining pipeline that processes unstructured clinical data in order to recognize clinical and social parameters such asage, gender, history of alcohol use, and violent behaviors, and explored the associations between these parameters and violent behaviors of patients with psychiatric disorders. The aim of our work was to demonstrate the feasibility of mining factors that are strongly associated with violent behaviors among psychiatric patients from unstructured psychiatric evaluation records using clinical text mining. Experiment results showed that stimulants, followed by a family history of violent behavior, suicidal behaviors, and financial stress were strongly associated with violent behaviors. Key aspects explicated in this paper include employing our text mining pipeline to extract clinical and social factors linked with violent behaviors, generating association rules to uncover possible associations between these factors and violent behaviors, and lastly the ranking of top rules associated with violent behaviors using statistical analysis and interpretation.

Network Security Events Analyze Method Based on Neural Networks and Genetic Algorithm

In view of the threat of network and information system, a network security events analyze method based on neural network and genetic algorithm is proposed in this paper. The neural network model is used to classify amounts of security events generated by various attack scenarios. And the rule items are extracted from the classification results. According to these items, the correlation rules are generated automatically by using genetic algorithm. Experiment results show that this method can classify events and generate association rules in the absence of manual intervention. And this automatic and self-adaptive method improves the efficiency of system.

Associative Classification Framework for Cancer Microarray Data

Having good cancer classifiers are crucial in order to give the most effective and cost saving treatments for patients. Microarray is one of the vital tools in cancer studies, as it allows the discovery of gene expression patterns and promises better accuracy of cancer classification. This paper presents an associative classification framework for microarray data. The proposed framework combined the strength of both filter method and association rule mining. The experimental results showed that the selected gene subsets from generated association rules can improve the accuracy and interpretability of classifiers.

Proposition of a Balanced Cross Entropy Measure in Association Rule Mining

오늘날 정보통신기술의 발달과 모바일 기기 이용의 확산에 의해 데이터가 폭증하게 됨으로써 빅 데이터의 환경이 도래하였다. 데이터 마이닝은 빅 데이터 시대에 주목받고 있는 기술 중의 하나이며, 이 중에서 연관성 규칙은 여러 가지 연관성 평가 기준에 의해 트랜잭션 내의 항목 간의 관련성을 결정하는 기법이다. 이 기법은 흥미도 측도를 평가 기준으로 하여 트랜잭션에 포함되어 있는 항목간의 관련성을 나타내기 때문에 인터넷 쇼핑몰, 보건 및 의료, 제조업, 그리고 교육 분야 등 현업에서 많이 활용되고 있다. 연관성 평가 기준의 기본적인 흥미도 측도에는 지지도, 신뢰도, 향상도 등 여러 가지 측도들이 연구되어왔다. 일반적으로 연관성 규칙 마이닝에서는 사용자에 의해 정해진 지지도의 최소값을 만족하는 항목집합을 생성한 후 신뢰도의 최저값을 만족하는 것을 연관성 규칙으로 생성한다. 본 논문에서는 규칙의 정보량을 수량화한 측도인 교차 엔트로피를 변형한 균형화된 교차 엔트로피를 연관성 평가 기준의 측도로 제안하여 예제를 통해 이 측도의 연관성 평가 기준으로서의 유용성을 고찰하고자 한다.Today, with the development of information and communication technology and the spread of mobile devices, the data has increased and big data environment has arrived. Data mining is one of the most remarkable technologies in the big data era. One of them is the association rule, which is used to determine whether items are related by various evaluation criteria. This technique is used in many fields such as internet shopping mall, health and medical care, manufacturing, and education because it shows the relation between the two items by using the interestingness measures as the evaluation criteria. The basic interestingness measures of association threshold include support, confidence, and lift. Generally, association rule mining generates a set of items satisfying the minimum value of support determined by the user, and then generates association rules that satisfy the minimum value of confidence. In this paper, we propose balanced entropy as a measure of association evaluation criterion by modifying cross entropy, which is a quantitative measureof the amount of information in a rule, and examine the usefulness of this measure as an association evaluation criterion through some examples.

Generating API Call Rules from Version History and Stack Overflow Posts

Researchers have shown that related functions can be mined from groupings of functions found in the version history of a system. Our first contribution is to expand this approach to a community of applications and set of similar applications. Android developers use a set of application programming interface (API) calls when creating apps. These API calls are used in similar ways across multiple applications. By clustering co-changing API calls used by 230 Android apps across 12k versions, we are able to predict the API calls that individual app developers will use with an average precision of 75% and recall of 22%. When we make predictions from the same category of app, such as Finance, we attain precision and recall of 81% and 28%, respectively. Our second contribution can be characterized as “programmers who discussed these functions were also interested in these functions.” Informal discussions on Stack Overflow provide a rich source of information about related API calls as developers provide solutions to common problems. By grouping API calls contained in each positively voted answer posts, we are able to create rules that predict the calls that app developers will use in their own apps with an average precision of 66% and recall of 13%. For comparison purposes, we developed a baseline by clustering co-changing API calls for each individual app and generated association rules from them. The baseline predicts API calls used by app developers with a precision and recall of 36% and 23%, respectively.

Mining Negative Correlation Biclusters from Gene Expression Data using Generic Association Rules

A majority of existing biclustering algorithms for microarrays data focus only on extracting biclusters with positive correlations of genes. Nevertheless, biological studies show that a group of biologically significant genes may exhibit negative correlations. In this paper, we propose a new biclustering algorithm, called NBic-ARM (Negative Biclusters using Association Rule Mining). Based on Generic Association Rules, our algorithm identifies negatively-correlated genes. To assess NBic-ARM’s performance, we carried out exhaustive experiments on three real-life datasets. Our results prove NBic-ARM’s ability to identify statistically and biologically significant biclusters.

Detecting Crime Patterns from Swahili Newspapers Using Text Mining

The Tanzania Police Force, as many other law enforcement agencies in developing countries, relies mostly on manual, personal judgments, and other inadequate tools for analysis of data in its crime databases. This approach is inadequate and prone to errors. Moreover, research shows that more than half of all crimes committed in Tanzania are not reported to police and thus it is likely that they are not analysed by the police. In this study, we use text mining to extract crime patterns from sources of crime data outside police databases. In fact, we use four daily published Swahili newspapers. With the help of our developed patterns mining model we extracted several crimes reported in the newspapers, we mapped the distribution of the mined crimes country-wide, and with the use of FP-growth, we generated association rules between the mined crimes. Results from this study will contribute to crime detection and prevention strategies.

IMIDB: An Algorithm for Indexed Mining of Incremental Databases

AbstractAssociation rules provide important knowledge that can be extracted from transactional databases. Owing to the massive exchange of information nowadays, databases become dynamic and change rapidly and periodically: new transactions are added to the database and/or old transactions are updated or removed from the database. Incremental mining was introduced to overcome the problem of maintaining previously generated association rules in dynamic databases. In this paper, we propose an efficient algorithm (IMIDB) for incremental itemset mining in large databases. The algorithm utilizes the trie data structure for indexing dynamic database transactions. Performance comparison of the proposed algorithm to recently cited algorithms shows that a significant improvement of about two orders of magnitude is achieved by our algorithm. Also, the proposed algorithm exhibits linear scalability with respect to database size.

Semantic association rule mining in text using domain ontology

This paper reports a procedure for ontology-based association rule mining for knowledge extraction from text. Association rule mining (ARM) algorithms have the limitations of generating many non-interesting rules, huge number of discovered rules, and low algorithm performance. This research demonstrates a procedure for improving the performance of ARM in text mining by using domain ontology. A study context of Nigerian politics using news text from a Nigerian online newspaper was selected, and a methodology that combined natural language processing, ontology-based keywords extraction, and the modified Generating Association Rules based on Weighting (GARW) scheme was applied. The result revealed significant rule reduction in the number of generated rules, and produced rules, which are more semantically related to the problem context when compared to when ARM approaches that are not ontology-based is used. The study shows that domain ontology can improve the performance of ARM algorithms when dealing with unstructured textual data.

SPATIAL ANALYSIS OF EARTHQUAKES IN IRAQ USING STATISTICAL AND DATA MINING TECHNIQUES

Statistical and data mining techniques (DMTs) are applied to an earthquakes catalogue of Iraq to study the spatial distribution pattern of earthquakes over the period from 1900 to 2010. The employed techniques are Quadrant Account Analysis (QCA), Tree-clustering, k-means Clustering, Association rules, and Linear Regression. Results of QCA showed that the pattern of earthquake occurrence beneath Iraq was spatially clustered. According to results of application of tree-clustering, earthquakes were grouped into nine clusters depending on degree of similarity between events. Results K-means clustering confirmed results of tree-clustering. Application of association rules failed to generate association rules between the earthquakes parameters (location, depth and magnitude, ...etc.). A weak relationship between depth and magnitude was the result of application of linear regression.

Mining Association Rules: A Case Study on Benchmark Dense Data

<p class="Abstract">Data mining is the process of discovering knowledge and previously unknown pattern from large amount of data. The association rule mining (ARM) has been in trend where a new pattern analysis can be discovered to project for an important prediction about any issues. Since the first introduction of frequent itemset mining, it has received a major attention among researchers and various efficient and sophisticated algorithms have been proposed to do frequent itemset mining. Among the best-known algorithms are Apriori and FP-Growth. In this paper, we explore these algorithms and comparing their results in generating association rules based on benchmark dense datasets. The datasets are taken from frequent itemset mining data repository. The two algorithms are implemented in Rapid Miner 5.3.007 and the performance results are shown as comparison. FP-Growth is found to be better algorithm when encountering the support-confidence framework.</p>

Enhancement of Effective Spatial Data Analysis using R

Background: The availability of Spatial Data which is a part of GIS is growing day by day in an exponential manner. This high availability of data is throwing challenges to the research community to analyze and draw effective conclusions. The Present Study aims at requirement for effective analysis and to draw Conclusions. Methods/Statistical Analysis: Spatial Analysis requires logical relationships between attribute data and map features.Spatial data Analysis is not a simple single task it requires complex procedures in which combinational techniques namely Hybrid techniques are required for effective analysis. Mathematics and statistics are the fundamentals to spatial data analytics. In this paper, a realistic Spatial crime data set was considered for analysis. It involves different types of data mining Techniques like Clustering, Classification and Association rule mining techniques apart from Hybrid techniques. These hybrid Data mining Techniques were applied using R. Findings: The Hybrid Data Mining techniques with K-means Clustering and J48 Decision Tree Algorithm was developed and Applied for the enhancement of accuracy. Association Rule generation Apriori algorithm was applied on the resultant K-means clustered data set. The application of 3D visualization techniques also made for further analysis. Applications/Improvements: It is essentially required to analyze these complex spatial data sets effectively. So there is a need of hybrid Spatial Data Mining Techniques requirement for effective analysis and to draw Conclusions.

Knowledge Discovery in Text Mining using Association Rule Extraction

Internet and information technology are the platform where huge amount of information is available to use. But searching the exact information for some knowledge is time consuming and results confusion in dealing with it. Retrieving knowledge manually from collection of web documents and database may cause to miss the track for user. Text mining is helpful to user to find accurate information or knowledge discovery and features in the text documents. Thus there is need to develop text mining approach which clearly guides the user about what is important information and what is not, how to deal with important information, how to generate knowledge etc. Knowledge discovery is an increasing field in the research. For a user reading the collection of documents and get some knowledge is time consuming and less effective. There has been a significant improvement in the research related to generating Knowledge Discovery from collection of documents. We propose a method of generating Knowledge Discovery in Text mining using Association Rule Extraction. Using this approach the users are able to find accurate and important knowledge from the collection of web documents which will reduce time for reading all those documents. General Terms Association Rule, Text mining

Research of Data-Aiming Mining Algorithm in Cloud Environment

Cloud computing contains a huge amount of data, which are featured as being widely distributed, heterogeneous, and dynamic. Thus, aiming at how to mine useful parts in these information, this paper proposes an Apriori algorithm based on cloud computing and introduces cost-sensitive learning and non-filter matrix to find k frequency set and uses the method of generating association rules to improve effectiveness of data mining. Simulation experiments show that mining algorithm in this paper is highly effective and suitable for data mining in the context of cloud computing. With the emergence of the concept cloud computing, more and more information is shared and spread through the Internet. How can carry out data mining in cloud computing became an important means of access to information (1). Literature (2) data mining solutions in the cloud computing environment is proposed, through the cloud computing and cloud computing services, describes the problem solving mechanism for data mining services. Literature (3) proposed a data processing method based on cloud computing, based on mining user browsing preference path. Experiments show that the mining algorithm for large amount of data logs, accuracy and efficiency than ordinary users preferred path mining algorithm based on frequency increased. Literature (4) proposed a new framework for data mining in the field of cloud platforms and cloud computing environments are described mechanism of data mining in the field of service-oriented. Literature (5) is a method of data mining based on cloud computing technology: large data sets and mining tasks on multiple computers in parallel, experiments show that use cloud computing technology to handle large data sets in a cluster, you can significantly improve the efficiency of data mining. Literature (6) for cloud computing resources prediction model based on data mining technology, results showed that this model not only improves the prediction accuracy of cloud computing resources, and reduce the complexity of modeling, improves the efficiency of models, to cloud computing resources provides a new way of building models. Literature (7) proposed data mining services view, cloud services task force has analyzed data mining advantages and challenges, data mining and prediction of cloud point data mining development trend of cloud services, providing data mining services framework for cloud services. This paper proposes an Apriori algorithm based on cloud computing and introduces cost-sensitive learning and non-filter matrix to find k

Boosting association rule mining in large datasets via Gibbs sampling

Current algorithms for association rule mining from transaction data are mostly deterministic and enumerative. They can be computationally intractable even for mining a dataset containing just a few hundred transaction items, if no action is taken to constrain the search space. In this paper, we develop a Gibbs-sampling-induced stochastic search procedure to randomly sample association rules from the itemset space, and perform rule mining from the reduced transaction dataset generated by the sample. Also a general rule importance measure is proposed to direct the stochastic search so that, as a result of the randomly generated association rules constituting an ergodic Markov chain, the overall most important rules in the itemset space can be uncovered from the reduced dataset with probability 1 in the limit. In the simulation study and a real genomic data example, we show how to boost association rule mining by an integrated use of the stochastic search and the Apriori algorithm.

An integrated method of associative classification and neuro-fuzzy approach for effective mammographic classification

Computer-aided diagnosis has gained a significant attention in helping radiologists to improve the accuracy of mammographic detection and diagnostic decision. The aim of proposed research work is to build an efficient and accurate classifier for the classification of mammogram images using a hybrid method by incorporating Genetic Association Rule Miner (GARM) with the learning capability of neural network. A set of features is extracted that comprises of 34 features based on the second and third level of the wavelet decomposition with 13 features measured directly from the gray-level co-occurrence matrix. In order to eliminate the inappropriate features and to increase the efficiency mining process, a multivariate filter is used for feature selection. Based on the selected features, an association rule mining based on modified GARM is used to generate association rules. In the classification phase, the newly generated association rules are used as the input for the creation and training of an artificial neural network. Furthermore, an extended associative classifier using fuzzy feed-forward backpropagation neural network (ACFNN) is proposed as an effective classifier in the context of mammography. The proposed ACFNN performance is compared with associative classifier using feed-forward backpropagation neural network (ACNN). Based on the experimental results, the performance of the proposed ACFNN is improved significantly. Furthermore, it can be inferred that the mammogram classification is better by using ACFNN with accuracy of 95.1 % as compared to ACNN with 93.7 %.

Novel Hybrid k-D-Apriori Algorithm for Web Usage Mining

The Web usage mining is a branch of web mining in which by clustering the datasets, frequently accessed patterns can be obtained for betterment of social portals, websites. The divisive analysis is one of the types of hierarchical method of data clustering that is used to separate each dataset from the clustered data depending on previous small clusters. Apriori & k-Apriori are commonly used hierarchical algorithms for web usage mining but they are less efficient for mining dynamic item sets such as twitter dataset. Recently D-Apriori algorithm has been proposed in the literature for mining dynamic dataset to find the frequently accessed web pages from web log database. The disadvantage of D-Apriori is that, it takes more execution time compared to k-Apriori. This research work proposes a new hybrid k-D-Apriori algorithm that reduces the execution time, improves frequent pattern generation, works efficiently with dynamic datasets and gives improved association rule generation as compared to D-Apriori algorithm.