Search

This study presents a novel approach for discovering actionable knowledge and exploring data-based models from data recorded by household smart meters. The proposed framework is supported by a machine learning architecture based on the application of data mining methods and spatial analysis to extract temporal and spatial restricted clusters of characteristic monthly electricity load profiles. In addition, it uses these clusters to perform short-term load forecasting (1 week) using recurrent neural networks. The approach analyses a database with measurements of 1000 smart meters gathered during 4 years in Guayaquil, Ecuador. Results of the proposed methodology led us to obtain a precise and efficient stratification of typical consumption patterns and to extract neighbour information to improve the performance of residential energy consumption forecasting The University of Girona and SENESCYT-Ecuador awarded the author with a pre-doctoral grant of Secretaría Nacional de Educación Superior, Ciencia, Tecnología e Innovación, (SENESCYT)— Ecuador. This work has been partially funded by the grant PID2020-117171RA-I00 funded by MCIN/AEI/10.13039/501100011033, the Government of Catalonia under 2017SGR1551 and the E-LAND project which received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 824388

Online privacy policies are lengthy and hard to comprehend. To address this problem, researchers have utilized machine learning (ML) to devise tools that automatically summarize online privacy policies for web users. One such tool is our free and publicly available browser extension, PrivacyCheck. In this paper, we enhance PrivacyCheck by adding a competitor analysis component–a part of PrivacyCheck that recommends other organizations in the same market sector with better privacy policies. We also monitored the usage patterns of about a thousand actual PrivacyCheck users, the first work to track the usage and traffic of an ML-based privacy analysis tool. Results show: (1) there is a good number of privacy policy URLs checked repeatedly by the user base; (2) the users are particularly interested in privacy policies of software services; and (3) PrivacyCheck increased the number of times a user consults privacy policies by 80%. Our work demonstrates the potential of ML-based privacy analysis tools and also sheds light on how these tools are used in practice to give users actionable knowledge they can use to pro-actively protect their privacy.

In recent years, high-throughput sequencing technologies provide unprecedented opportunity to depict cancer samples at multiple molecular levels. The integration and analysis of these multi-omics datasets is a crucial and critical step to gain actionable knowledge in a precision medicine framework.This paper explores recent data-driven methodologies that have been developed and applied to respond major challenges of stratified medicine in oncology, including patients’ phenotyping, biomarker discovery and drug repurposing. We systematically retrieved peer-reviewed journals published from 2014 to 2019, select and thoroughly describe the tools presenting the most promising innovations regarding the integration of heterogeneous data, the machine learning methodologies that successfully tackled the complexity of multi-omics data, and the frameworks to deliver actionable results for clinical practice. The review is organized according to the applied methods: Deep learning, Network-based methods, Clustering, Features Extraction and Transformation, Factorization. We provide an overview of the tools available in each methodological group and underline the relationship among the different categories.Our analysis revealed how multi-omics datasets could be exploited to drive precision oncology, but also current limitations in the development of multi-omics data integration.

Actionable Knowledge Discovery approaches to extract the business and technical significant actions/patterns to support direct decision making. These actions suggest how to transform an object from an undesirable status to a desirable status by incurring less cost and high profit. This article aims to propose a work that generates actionable patterns efficiently. It reduces the search space and number of iterations for attribute value change during action generation. Performance of the proposed method is compared with Yang's method and OF-CEAMA on the basis of four parameters i.e. the total number of rules required for action generation, run time of the methods, the total number of generated actions, total net profit and time and space complexity. Experiments have been carried out on four datasets retrieved from UCI Machine learning repository. Experimental results show that the proposed work takes less time than the other two methods to extract actions for all datasets. Also, the number of rules required to generate actions are less than the other two methods. Results also suggest that a decrease in execution time does not compromise the information and proposed work generates the same actions and net profit. Moreover, the proposed work tries to transfer an object from undesired status to the desired status.

A lack of alignment between professional practice and education has triggered the move to alternative educational models, often with detailed scripts and templates to be followed. Among these are variants of problem/case-based models, where learners are challenged to achieve professionally desired learning outcomes and acquire knowledge and skills in their respective disciplines by encountering real-life situations as the stimulus and focus of their learning activities. A characteristic of these diverse models is that their scripts have been based on theories and understandings of what constitutes good learning and teaching in general. This article reports on a study conducted among law students that uses a script which focuses on the core ‘know-how’ of the legal profession. To examine the merits of this approach with respect to actionable knowledge, we employ perspectives from Knorr Cetina’s practice-theoretical lens. Particular attention is paid to how students are introduced to and learn the three archetypes of epistemic practices that Knorr Cetina identifies as central for members of expert cultures. The results from the study not only show that these practices were developed but also detail what is important in this respect. Thus, the article addresses calls in this journal for more research on how connections between school and work for professionals can be enhanced.

Although progress has been made to advance our understanding of the risks involved in offshore oil extraction activities, a regional scale understanding of factors contributing to losses in infrastructure integrity are lacking. Recent data integration efforts have resulted in a comprehensive database that allows for an unprecedented study of the external and internal factors that impact the structural and operational integrity of offshore platforms in the Gulf of Mexico. This study constitutes some of the initial explorations into that database by focusing on the relationships among a diverse set of variables and the age at which a platform is removed. We apply Geographically Weighted Regression to account for the heterogeneity of the operating environment, finding robust yet unexpected relationships that shed light on some of the factors that influence platform removal. Our findings pave the way for future studies aimed at building actionable knowledge.

Data mining is used to extract actionable knowledge from huge amount of raw data. In numerous real life applications, data are stored in sequential form, hence mining sequential patterns has been one of the most popular fields in data mining. Due to its various applications, across the past decades, a significant number of literature have addressed this problem and provided elegant solutions. In this paper we propose a novel tree data structure, SP-Tree, to store the sequence database in a new and efficient manner. Additionally, we propose a new mining algorithm Tree-miner to mine sequential patterns from SP-Tree. To further enhance the performance of our algorithm, we incorporate multiple pruning techniques and optimizations. As our SP-Tree stores the complete database, it can also be used for incremental and dynamic databases, tree-structure is particularly advantageous for interactive mining. We demonstrate how our SP-Tree based Tree-miner algorithm significantly outperforms all of the existing state-of-the-art algorithms, across 6 real life datasets. We conclude by discussing the possible extensions of our approach to other related fields of sequential data.

What should we teach students who are interested in geospatial data science and what should an undergraduate or graduate curriculum in this area look like? This paper addresses such issues from an epistemological perspective and discusses the critical linkages among different fields that are related to geo computational thinking.

Local changes in the structure of G-protein coupled receptors (GPCR) binders largely affect their pharmacological profile. While the sought efficacy can be empirically obtained by introducing local modifications, the underlining structural explanation can remain elusive. Here, molecular dynamics (MD) simulations of the eticlopride-bound inactive state of the Dopamine D3 Receptor (D3DR) have been clustered using a machine learning-based approach in the attempt to rationalize the efficacy change in four congeneric modulators. Accumulating extended MD trajectories of receptor-ligand complexes, we observed how the increase in ligand flexibility progressively destabilized the crystal structure of the inactivated receptor. To prospectively validate this model, a partial agonist was rationally designed based on structural insights and computational modeling, and eventually synthesized and tested. Results turned out to be in line with the predictions. This case study suggests that the investigation of ligand flexibility in the framework of extended MD simulations can assist and inform drug design strategies, highlighting its potential role as a powerful in silico counterpart to functional assays.

Water quality measures for inland and coastal waters are available as discrete samples from professional and volunteer water quality monitoring programs and higher-frequency, near-continuous data from automated in situ sensors. Water quality parameters also are estimated from model outputs and remote sensing. The integration of these data, via data assimilation, can result in a more holistic characterization of these highly dynamic ecosystems, and consequently improve water resource management. It is becoming common to see combinations of these data applied to answer relevant scientific questions. Yet, methods for scaling water quality data across regions and beyond, to provide actionable knowledge for stakeholders, have emerged only recently, particularly with the availability of satellite data now providing global coverage at high spatial resolution. In this paper, data sources and existing data integration frameworks are reviewed to give an overview of the present status and identify the gaps in existing frameworks. We propose an integration framework to provide information to user communities through the the Group on Earth Observations (GEO) AquaWatch Initiative. This aims to develop and build the global capacity and utility of water quality data, products, and information to support equitable and inclusive access for water resource management, policy and decision making.

In this study, an understanding and a review of Knowledge Discovery Database (KDD) development and its applications in tire maintenance are highlighted. Even though data mining has been successful in becoming a major component of various business processes and applications, the benefits and real-world expectations are very important to consider. It is also surprising to note that very little is known to date about the usefulness of applying knowledge discovery in transport related research. From the literature, the frameworks for carrying out knowledge discovery and data mining have been revised over the years to meet the business requirements. The Domain Driven Data Mining (DDDM) is one of the KDD frameworks often used for this purpose. In this study, we apply DDDM-KDD for formulating effective tire maintenance strategy within the context of a Malaysian's logistics company. We also discussed the weaknesses of the results from DDDM-KDD and emphasize the important of using the next generation of KDD framework Actionable Knowledge Discovery (AKD) for an effective decision. The direction flow of research, research methods use and contribution of research also are highlighted.

Actionable Knowledge Discovery has attracted much interest lately. It is almost a new paradigm shift toward mining more usable and more applicable knowledge in each specific domain. An action is a new tool in this research area that suggests some changes to the user to gain a profit in his/her domain. Currently, most of action mining methods rely on simple data which describes each object independently. Since social data has more complex structure due to the relationships between individuals, a major problem is that such structural information is not taken into account in the action mining process. This leads to miss some useful knowledge and profitable actions. Consequently, more effective methods are needed for mining actions. The main focus of this work is to extract cost-effective actions from social networks in which nodes have attributes. The actions suggest optimal changes in nodes’ attributes that are likely to result in changing labels of users to more desired one when they are applied. We develop an action mining method based on Random Walks that naturally combines the information from the network structure with nodes attributes. We formulate action mining as an optimization problem where the goal is to learn a function that varies the values of nodes’ attributes which in turn affect edges’ weights in the network so that the labels of intended individuals are likely to take the desired label while minimizing the cost of incurring the changes. Experiments confirm that the proposed approach outperforms the current state-of-the-art in action mining.

With healthcare fraud accounting for financial losses of billions of dollars each year in the United States, the task of investigating regulation adherence is key to reduce the impact of Fraud, Waste and Abuse (FWA) on the healthcare industry. Providers rendering services to patients typically submit claims to healthcare insurance agencies. Such claims must follow specific compliance criteria specified by state and federal policies. This paper presents an ontology-based system that aims to support the FWA claim investigation process by extracting graph-based actionable knowledge from policy text describing those compliance criteria. We discuss the process of creating a domain-specific ontology to model human experts’ conceptualisations and to incorporate early-on the feedback of FWA investigators, who are the early adopters of our solution. We explore whether the ontology is expressive and flexible enough to model the diverse compliance processes and complex relationships defined in policy documents. The ontology is then used, in combination with natural language understanding and semantic techniques, to guide the extraction of a Knowledge Graph (KG) from policies. Our solution is validated in terms of correctness and completeness by comparing the extracted knowledge to a ground truth created by investigators. Lastly, we discuss further challenges our deployed semantic system needs to tackle in this novel scenario, with the prospect of supporting the investigation process.

Conventional hypothesis testing is carried out in a hypothesis-driven manner. A scientist must first formulate a hypothesis based on what he or she sees and then devise a variety of experiments to test it. Given the rapid growth of data, it has become virtually impossible for a person to manually inspect all data to find all of the interesting hypotheses for testing. In this article, we propose and develop a data-driven framework for automatic hypothesis testing and analysis. We define a hypothesis as a comparison between two or more subpopulations. We find subpopulations for comparison using frequent pattern mining techniques and then pair them up for statistical hypothesis testing. We also generate additional information for further analysis of the hypotheses that are deemed significant. The number of hypotheses generated can be very large, and many of them are very similar. We develop algorithms to remove redundant hypotheses and present a succinct set of significant hypotheses to users. We conducted a set of experiments to show the efficiency and effectiveness of the proposed algorithms. The results show that our system can help users (1) identify significant hypotheses efficiently, (2) isolate the reasons behind significant hypotheses efficiently, and (3) find confounding factors that form Simpson’s paradoxes with discovered significant hypotheses.

In today’s world there is a need of speedy tools and techniques to convert big data into information. Traditional techniques are robust but might carry inherent bias of data scientist’s modeling skills. Machine Learning (ML) models makes the machine learn from the data and might not suffer from human bias. Some researchers argue that ML techniques can turn around model faster for big data, while others debate traditional techniques would still outperform when data is not so big. This paper tests arguments raised above in Oil & Gas (ONG) industry for actionable knowledge extraction on not so big sample customer data to predict customer’s campaign response. Our experiment reveals that on our data ML doesn’t outclass traditional modeling but is either slightly better or at par in terms of model accuracy i.e. percentage of instances classified correctly. It also establishes that though ML results don’t improve much on accuracy but some of them can be developed much faster.

Healthcare spending has been increasing in the last few decades. One of the main reasons for this increase is hospital readmissions, which is defined as a re-hospitalization of a patient after being discharged from a hospital within a short period of time. The excessive amount of money spent every year on hospital readmissions and the urge to enhance healthcare quality make reducing hospital readmissions a necessity. In this paper, we extract knowledge from a medical dataset and apply the concept of mining actionable rules to guide the health domain experts in their decision-making process. We present novel algorithms to increase the predictability of the patients' paths (the sequence of procedures that patients undertakes to reach a desired treatment) by clustering the patients according to their set of diagnoses. Moreover, we present a scoring metric to evaluate procedures in procedure graphs (the tree of all possible procedure paths) and a scoring metric to evaluate clusters of diagnoses which would allow us to anticipate the number of following readmissions for a new patient. Finally, we present an algorithm to evaluate the score (average number of following readmissions) for new patients prior to applying the action rules and after. The results presented in this paper show that our algorithms are able to reduce the average number of readmissions to a high degree.

The majority of the conventional mining algorithms treat the mining process as an isolated data-driven procedure and overlook the semantic of the targeted data. As a result, the generated patterns are abundant and end users cannot act upon them seamlessly. Furthermore, interdisciplinary knowledge can not be obtained from domain-specific silo of data. The emergence of Linked Data (LD) as a new model for knowledge representation, which intertwines data with its semantics, has introduced new opportunities for data miners. Accordingly, this paper proposes an ontology-based Semantic-Aware Bayesian network (BN) model. In contrast to the existing mining algorithms, the proposed model does into transform the original format of the LD set. Therefore, it not only accommodates the semantic aspects in LD, but also caters to the need of connecting different data-sets from different domains. We evaluate the proposed model on a Bone Dysplasia dataset, Experimental results show promising performance.

Actionable Knowledge Discovery has attracted much interest lately. It is almost a new paradigm shift toward mining more usable and more applicable knowledge in each specific domain. Action Rule is a new tool in this research area that suggests some actions to user to gain a profit in his/her domain. Up to now some methods have been devised for action rule mining. Decision Trees, Classification Rules and Association Rules are three learner machines that already have been used for action rule mining. But when we want to suggest an action we need to know the causal relationships among parameters and current methods can’t say anything about that. So that we use here Bayesian Networks as one of the most powerful knowledge representing models that can show the causal relationships between variables of interest for extracting action rules. Another benefit of new method is about the background knowledge. Bayesian Networks are very powerful at integrating the background knowledge into model. At the end of this paper an action rule mining system is proposed that can suggest the most profitable action rules for each case or class of cases.

In this work, we define procedure paths as the sequence of procedures that a given patient undertakes to reach a desired treatment. In addition to its value as a mean to inform the patient of his or her course of treatment, being able to identify and anticipate procedure paths for new patients is an essential task for examining and evaluating the entire course of treatments in advance, and ultimately rectifying undesired procedure paths accordingly. In this paper, we first introduce two approaches for anticipating the state of the patient that he or she will end up in after performing some procedure p; the state of the patient will consequently indicate the following procedure that the patient is most likely to undergo. By clustering patients into subgroups that exhibit similar properties, we improve the predictability of their procedure paths, which we evaluate by calculating the entropy to measure the level of predictability of following procedure. The clustering approach used is essentially a way of personalizing patients according to their properties. The approach used in this work is entirely novel and was designed specifically to address the twofold problem of first being able to predict following procedures for new patients with high accuracy, and secondly being able to construct such groupings in a way that allows us to identify exactly what it means to transition from one cluster to another. Then, we further devise a metric system that will evaluate the level of desirability for procedures along procedure paths, which we would subsequently map to a metric system for the extracted clusters. This will allow us to find desired transitions between patients in clusters, which would result in reducing the number of anticipated readmissions for new patients.

Incident reporting enables clinicians to examine historical patient safety events and to target different levels of analysis toward actionable knowledge. The cross-cultural adaptation of reporting instruments promotes the international communication on medical errors and patient safety culture.This study initializes a translation and adaptation of the Common Formats (in US) to Chinese and a localized reporting on perinatal incidents in a Chinese hospital. The results demonstrate the validity of the cross-cultural translation and diversity in a typical perinatal incident reported by Chinese clinicians. These findings suggest (1) a comprehensive data report format is critical in the incident reporting; (2) an imperative need of cross-cultural study on incident reporting; (3) future direction of incident reporting and patient safety culture.

Additive tree models (ATMs) are widely used for data mining and machine learning. Important examples of ATMs include random forest, adaboost (with decision trees as weak learners), and gradient boosted trees, and they are often referred to as the best off-the-shelf classifiers. Though capable of attaining high accuracy, ATMs are not well interpretable in the sense that they do not provide actionable knowledge for a given instance. This greatly limits the potential of ATMs on many applications such as medical prediction and business intelligence, where practitioners need suggestions on actions that can lead to desirable outcomes with minimum costs.To address this problem, we present a novel framework to post-process any ATM classifier to extract an optimal actionable plan that can change a given input to a desired class with a minimum cost. In particular, we prove the NP-hardness of the optimal action extraction problem for ATMs and formulate this problem in an integer linear programming formulation which can be efficiently solved by existing packages. We also empirically demonstrate the effectiveness of the proposed framework by conducting comprehensive experiments on challenging real-world datasets.

Among the most important and distinctive actionable knowledge are actionable behavioral rules that can directly and explicitly suggest specific actions to take to influence (restrain or encourage) the behavior in the users' best interest. The problem of mining such rules is a search problem in a framework of support and expected utility. The previous definition of a rule's support assumes that each instance which supports a rule has the uniform contribution to the support. However, this assumption is usually violated in practice to some extent, and thus will hinder the performance of algorithms for mining such rules. In this paper, to handle this problem, a power-function-based observation-weighting model for support and corresponding mining algorithm are proposed. The experimental results strongly suggest the validity and the superiority of our approach.

Argumentation-based multi-agent collaborative classification is a promising paradigm for reaching agreements in distributed environments. In this paper, we advance the research by introducing a new domain ontology enriched inductive learning approach for collaborative classification, in which agents are able to constructing arguments taking into account their own domain knowledge. This paper focuses on classification rules inductive learning, and presents Arguing SATE-Prism, a domain ontology enriched approach for multi-agent collaborative classification based on argumentation. Domain ontology, in this context, is exploited for driving a paradigm shift from traditional data-centered hidden pattern mining to domain-driven actionable knowledge discovery. Preliminary experimental results show that higher classification accuracy can be achieved by exploiting ontological reasoning in argumentation based multi-agent collaborative classification. Our experiments also demonstrate that the proposed approach out-performs comparable classification paradigms in presence of instances with missing values, harnessing the advantages offered by ontological reasoning.

Stream data has become ubiquitous due to advances in acquisition technology and pervades numerous applications. These massive data gathered as continuous flow, are often accompanied by dire need for real-time processing. One aspect of data streams deals with storage management and processing of continuous queries for aggregation. Another significant aspect pertains to discovery and understanding of hidden patterns to derive actionable knowledge using mining approaches. This chapter focuses on stream clustering and presents a primer of clustering algorithms in data stream environment.

Prescribed treatments to patients often result in side effects that may not be known beforehand. Side effects analysis research focuses on specific treatments and targets small groups of patients. In previous work, we presented methods for extracting treatment effects from the Florida State Inpatient Databases (SID), which contain over 2.5 million visit discharges from 1.5 million patients. We classified these effects into positive, neutral, and negative effects. In addition, we proposed an approach for clustering patients based on negative side effects and analyzed them. As an extension to this work, We believe that a system identifying the cluster membership of a patient prior to applying the procedure is highly beneficial. In this paper, we extended our work and introduced a new approach for predicting patients' negative side effects before applying a given meta-action (or procedure). We propose a system that measures the similarity of a new patient to existing clusters, and makes a personalized decision on the patient's most likely negative side effects. We further evaluate our system using SID, which is part of the Healthcare Cost and Utilization Project (HCUP). Our experiments validated our approach and produced desired results.

In this paper, we design and implement a novel system for connecting cyber threat information. The objective is to improve the information and its analysis results with machine intelligence assisted by human intelligence. This paper illustrates the system named NECOMAtter based on these assumptions, and summarizes our contributions in order to develop actionable knowledge.

Knowledge is considered actionable if users can take direct actions based on such knowledge to their advantage. Among the most important and distinctive actionable knowledge are actionable behavioral rules that can directly and explicitly suggest specific actions to take to influence (restrain or encourage) the behavior in the users’ best interest. However, in mining such rules, it often occurs that different rules may suggest the same actions with different expected utilities, which we call conflicting rules. To resolve the conflicts, a previous valid method was proposed. However, inconsistency of the measure for rule evaluating may hinder its performance. To overcome this problem, we develop a new method that utilizes rule ranking procedure as the basis for selecting the rule with the highest utility prediction accuracy. More specifically, we propose an integrative measure, which combines the measures of the support and antecedent length, to evaluate the utility prediction accuracies of conflicting rules. We also introduce a tunable weight parameter to allow the flexibility of integration. We conduct several experiments to test our proposed approach and evaluate the sensitivity of the weight parameter. Empirical results indicate that our approach outperforms those from previous research.

Energy consumption of mobile applications is nowadays a hot topic, given the widespread use of mobile devices. The high demand for features and improved user experience, given the available powerful hardware, tend to increase the apps’ energy consumption. However, excessive energy consumption in mobile apps could also be a consequence of energy greedy hardware, bad programming practices, or particular API usage patterns. We present the largest to date quantitative and qualitative empirical investigation into the categories of API calls and usage patterns that—in the context of the Android development framework—exhibit particularly high energy consumption profiles. By using a hardware power monitor, we measure energy consumption of method calls when executing typical usage scenarios in 55 mobile apps from different domains. Based on the collected data, we mine and analyze energy-greedy APIs and usage patterns. We zoom in and discuss the cases where either the anomalous energy consumption is unavoidable or where it is due to suboptimal usage or choice of APIs. Finally, we synthesize our findings into actionable knowledge and recipes for developers on how to reduce energy consumption while using certain categories of Android APIs and patterns

Among the most important and distinctive actionable knowledge are actionable behavioral rules that can directly and explicitly suggest specific actions to take to influence the behavior in the users' best interest. However, in mining such rules, it often occurs that different rules may suggest the same actions with different expected utilities, which we call conflicting rules. To resolve the conflicts, a previous pruning method was proposed. However, inconsistency of the measure for rule pruning may hinder its performance. To overcome this problem, we develop a new pruning method to achieve rule pruning in actionable rule discovery. We conduct several experiments to test our proposed approach and evaluate the sensitivity of the weight parameter. Empirical results based on a benchmark terrorism dataset indicate that our approach outperforms those from previous research.

Big Data promises to transform the rapidly expanding collection of health data we see today into smart, relevant, and actionable knowledge to deliver personalized health. Yet, traditional methods of clinical research cannot handle the volume, variability, velocity, and veracity of data that is currently being generated. Successful, real-world examples of this Big Data in health, therefore, are rare, and important problems remain unsolved. Not surprisingly, some now question if the ultimate impact of Big Data in medicine will be largely incremental rather than transformational. This is a story that needs to be changed, and …

Most of the mining techniques has only concerned with interesting patterns. However, in the recent years, there is an increasing demand of mining the Unexpected Items or Outliers or Rare Items. Several application domains have realized the direct mapping between outliers in data and real world anomalies that are of great interest to an analyst. Outliers represents semantically correct but infrequent situation in a database. Detecting outliers allows extracting useful and actionable knowledge to the domain experts. Our method which is used to extract temporal outliers is based on association rules, to infer the normal behavior of objects by extracting frequent rules from Stock Market dataset which is of quantitative, multidimensional and time series in nature. Then the temporal association rules are pruned by the unexpectedness measure - Residual Leverage to generate Temporal Outliers. The temporal outliers from Stock Market database are the Stock Split that occurs on a particular day. A stock split results in a stock price increase following the decrease immediately. This split provides a signal to the market which in turn increases the demand and prices.

Data mining techniques have been successfully applied in stock, insurance, medicine, banking and retailing domains. In the sport domain, for transforming sport data into actionable knowledge, coaches can use data mining techniques to plan training sessions more effectively, and to reduce the impact of testing activity on athletes. This paper presents one such model, which uses clustering techniques, such as improved K-Means, Expectation-Maximization (EM), DBSCAN, COBWEB and hierarchical clustering approaches to analyze sport physiological data collected during incremental tests. Through analyzing the progress of a test session, the authors assign the tested athlete to a group of athletes and evaluate these groups to support the planning of training sessions.

Real-world problems often call for efficient methods to discovery actionable knowledge on which business can directly act [3]. Some works for discovering actionable knowledge [3,5] view actions as behaviors which render a state of an instance into a preferred state, where a state is represented by feature values of the instance and whether a state is preferred is determined by a classifier. Actions usually have many-to-many relations with features of an instance. That is, an action may affect multiple features of an instance, and vise versa, a feature may be influenced by multiple actions. This type of many-to-many relationships between actions and features is prevalent in real-world applications. However, most existing works [3,5] only deal with one-to-one relationship and ignore manyto- many relationship between actions and features. In these works, an action is treated as a behavior with a fixed execution cost. Restricting to a one-to-one relationship between actions and features may not yield an action set (i.e. a set of actions) with the minimal total execution cost. Moreover, one-to-one relationship is simply a special case of many-to-many relationship, and hence the latter will be applicable to more real-world problems. Therefore we aim to extract action sets from a classifier for which the total execution cost is minimal based on many-to-many relationship between actions and features.

Background As a growing and more diverse amount of genetic and molecular profile data are generated in several models of Huntington’s disease (HD), the potential of these datasets for enhancing HD research has significantly increased. Network-based analysis has the capacity to identify robust HD-associated rules and signatures from complex datasets. These rules and signatures can in turn be exploited for better understanding HD biology on a global scale as well as for prioritising models and genes around questions of specific interest, thereby supporting decision making and biomedical intelligence in HD research. Aim and method To support biomedical intelligence in HD, and as part of the project and activities of the EHDN’s working group ‘Biological modifiers’, we have developed Biogemix, a network-based data integration framework for precise rule/pattern extraction across models and species. We used this framework for integrating greater than 16 publically available datasets including transcriptomic and gene perturbation data. Outcome We found previously suspected as well as novel HD-associated features, the latter providing new insights into the global effects of mutant huntingtin pathogenicity. Additionally, we managed to determine the functional distances between experimental models of HD and assess the biological relevance of individual models of HD, functional profiles of biological modifiers and shared/unshared properties between pathological and compensatory/survival genes. Conclusion Our results validate the value and precision of the Biogemix framework for actionable knowledge in HD. As larger HD datasets become available, our results suggest that applying the Biogemix approach to the analysis of these datasets will provide deep insights into the essential features of mutant huntingtin pathogenicity while providing strong guidance to translational and pharmacological research in HD. Support CHDI Foundation, ANR, EHDN, ANRT/GSK.

Incredible amounts of data exist across all domains of sports. This data can come in the form of individual player performance, coaching or managerial decisions, game-based events and/or how well the team functions together. The task is not how to collect the data, but what data should be collected and how to make the best use of it. By finding the right ways to make sense of data and turning it into actionable knowledge, sports organizations have the potential to secure a competitive advantage versus their peers. This knowledge seeking approach can be applied throughout the entire organization. From players improving their game-time performance using video analysis techniques, to scouts using statistical analysis and projection techniques to identify what talent will provide the biggest impact, data mining is quickly becoming an integral part of the sports decision making landscape where manager/coaches using machine learning and simulation techniques can find optimal strategies for an entire upcoming season.

The increased competition faced by today's companies can wield data mining tools to extract actionable knowledge and then use it as a weapon to outmaneuver competitors and boost revenue. Mining reclassification rules is a way to model actionable patterns directly from a given data set. The previous work on reclassification rule mining has shown that they are effective when variables are weakly correlated. However, when the data set is correlated, some redundant rules are in the result set. This problem becomes critical for discovering rules in correlated data which may have long frequent factor-sets. In this paper, we investigate properties of reclassification rules and offer a new method to discovery a set of non-redundant reclassification rules without information loss.

Data Mining for Business Applications presents the state-of-the-art research and development outcomes on methodologies, techniques, approaches and successful applications in the area. The contributions mark a paradigm shift from data-centered pattern mining to domain driven actionable knowledge discovery for next-generation KDD research and applications. The contents identify how KDD techniques can better contribute to critical domain problems in theory and practice, and strengthen business intelligence in complex enterprise applications. The volume also explores challenges and directions for future research and development in the dialogue between academia and business.

In recent years, multi-criteria optimization (MCO) community has made noticeable progress in the area of data mining and knowledge discovery. While most research effort is devoted to developing models and algorithms to "mine" data, not enough attention has been paid to the "knowledge discovery" aspect. Real-world data mining problems are complex and require close collaboration between data miners and domain experts. This paper analyzes the characteristics of MCO methods and proposes a framework that supports the integration of domain knowledge, business constraints and expectations, and data mining expertise. The aim of the framework is to turn the results of MCO-based data mining methods into actionable knowledge that can be applied to real-world problems.

Association mining often produces large collections of association rules that are difficult to understand and put into action. In this paper, we have designed a novel notion of combined patterns to extract useful and actionable knowledge from a large amount of learned rules. We also present definitions of combined patterns, design novel metrics to measure their interestingness and analyze the redundancy in combined patterns. Experimental results on real-life social security data demonstrate the effectiveness and potential of the proposed approach in extracting actionable knowledge from complex data.

GUI regression testing is challenging as GUIs are changed frequently during the whole SDLC. Currently, there are no effective techniques for doing GUI regression testing. In this paper, a novel actionable knowledge model (AKM) is proposed for representing GUI test cases. The proposed model is adaptive to the change of GUIs. A multi-agent based framework that infusing AKM for GUI application regression testing is developed to facilitate test case creation, execution and repair.

Automated all source data fusion primarily fuses analyst generated messages. The Automated Targeting Data Fusion (ATDF) effort created an application that rapidly associated very diverse information sources automatically. In recent military conflict situations, enormous amounts of multi-source data have been made available to war-fighters. These data include imagery, signals intelligence, acoustic information, input from human analysts, and other sources. The significant magnitude of collected intelligence overwhelms analysts, operators, collection managers and commanders. Complicating matters further is that the data is collected, transmitted, processed, exploited and disseminated via multiple stovepipe architectures involving different types of intelligence and support personnel. It is difficult for diverse users to maintain situational awareness over the battle-space. This is an extremely crucial issue with rapid targeting decisions. This ATDF capability offers an alternative to automatically support each individual user as well as decision makers who must use all of their collective command-level decisions. ATDF rapidly associates diverse intelligence products as they are collected and fuses it into actionable knowledge in near real-time.

The Advanced Agent Technology Laboratory in the School of Computer Science at Carnegie Mellon University, the Human Computer Interaction Group at the University of Pittsburgh, the Munitions Directorate of Air Force Research Laboratory (MN/AFRL), Rome Labs and Northrup Grumman are partnering in early-stage research to address issues in high-level information fusion, including intent inferencing and retasking of sensors and munitions. To help translate information superiority to decision superiority (i.e. to rapidly arrive at better decisions than adversaries can respond to), many scientific, technical and technological challenges must be addressed. The most critical of those are information fusion and management at different levels, communication, planning and execution monitoring. The RETSINA multiagent infrastructure allows information producers and users to discover one another and establish direct links. The robust, decentralized Infosphere which results, can be stood up rapidly and ensures that information of the specified types will be delivered to the right users under the right conditions. We will present the overall approach for this research program, the role and contribution of each of the partners, the envisioned testbed, examples and research results to-date. In addition, we will discuss plans for the out-years.

This chapter presents two methods that combine data mining and decision support techniques with the aim to generate actionable knowledge. Both methods follow the same methodology in which data mining is used to support decision-making. The methodology consists of the following phases: business understanding; data acquisition, data understanding and preprocessing; data mining through subgroup discovery; subgroup evaluation; and deployment for decision support. The two methods have been applied to support decisionmaking in marketing.

In response to attacks against enterprise networks, administrators increasingly deploy intrusion detection systems. These systems monitor hosts, networks, and other resources for signs of security violations. The use of intrusion detection has given rise to another difficult problem, namely the handling of a generally large number of alarms. In this paper, we mine historical alarms to learn how future alarms can be handled more efficiently. First, we investigate episode rules with respect to their suitability in this approach. We report the difficulties encountered and the unexpected insights gained. In addition, we introduce a new conceptual clustering technique, and use it in extensive experiments with real-world data to show that intrusion detection alarms can be handled efficiently by using previously mined knowledge.

This paper discusses actionable knowledge generation. Actionable knowledge is explicit symbolic knowledge, typically presented in the form of rules, that allows the decision maker to recognize some important relations and to perform an action, such as targeting a direct marketing campaign, or planning a population screening campaign aimed at targeting individuals with high disease risk. The disadvantages of using standard classification rule learning for this task are discussed, and a subgroup discovery approach proposed. This approach uses a novel definition of rule quality which is extensively discussed.

Traditionally, knowledge actionability has been investigated mainly by developing and improving technical interestingness. Recently, initial work on technical subjective interestingness and business-oriented profit mining presents general potential, while it is a long-term mission to bridge the gap between technical significance and business expectation. In this paper, we propose a two-way significance framework for measuring knowledge actionability, which highlights both technical interestingness and domain-specific expectations. We further develop a fuzzy interestingness aggregation mechanism to generate a ranked final pattern set balancing technical and business interests. Real-life data mining applications show the proposed knowledge actionability framework can complement technical interestingness while satisfy real user needs. © 2007, Inderscience Publishers.

Most data mining algorithms and tools stop at discovered customer models, producing distribution information on customer profiles. Such techniques, when applied to industrial problems such as customer relationship management (CRM), are useful in pointing out customers who are likely attritors and customers who are loyal, but they require human experts to postprocess the mined information manually. Most of the postprocessing techniques have been limited to producing visualization results and interestingness ranking, but they do not directly suggest actions that would lead to an increase the objective function such as profit. Here, we present a novel algorithm that suggest actions to change customers from an undesired status (such as attritors) to a desired one (such as loyal) while maximizing objective function: the expected net profit. We develop these algorithms under resource constraints that are abound in reality. The contribution of the work is in taking the output from an existing mature technique (decision trees, for example), and producing novel, actionable knowledge through automatic postprocessing.

This era of groundbreaking scientific developments in high-resolution, high-throughput technologies is allowing the cost-effective collection and analysis of huge, disparate datasets on individual health. Proper data mining and translation of the vast datasets into clinically actionable knowledge will require the application of clinical bioinformatics. These developments have triggered multiple national initiatives in precision medicine—a data-driven approach centering on the individual. However, clinical implementation of precision medicine poses numerous challenges. Foremost, precision medicine needs to be contrasted with the powerful and widely used practice of evidence-based medicine, which is informed by meta-analyses or group-centered studies from which mean recommendations are derived. This “one size fits all” approach can provide inadequate solutions for outliers. Such outliers, which are far from an oddity as all of us fall into this category for some traits, can be better managed using precision medicine. Here, we argue that it is necessary and possible to bridge between precision medicine and evidence-based medicine. This will require worldwide and responsible data sharing, as well as regularly updated training programs. We also discuss the challenges and opportunities for achieving clinical utility in precision medicine. We project that, through collection, analyses and sharing of standardized medically relevant data globally, evidence-based precision medicine will shift progressively from therapy to prevention, thus leading eventually to improved, clinician-to-patient communication, citizen-centered healthcare and sustained well-being.