User Approach to Knowledge Discovery in Networked Environment

Collaboration and networking demands are increasing and lots of organizational communicative activities have moved into technical networks. Need to understand not only how to refine right information contents out of the available data mass but also what type of information is important in various information using situations has increased. This chapter delves into the problem area of finding ways to support users to find relevant, specific types of information that is related to various phases of operating in network. Establishing a network, planning operations and managing operations differ from each others what comes into information requirements. It will be shown via four generalized cases that information requirements vary depending on what phase of networking activity the organization is. Via those cases that are based on sufficiently broad empirical material it will be cleared that knowledge requirements differ from situation to another. This leads to a conclusion that flexible data mining and knowledge discovery systems shall be constructed.

Data Mining and the Project Management Environment

The chapter illustrates how data mining and knowledge management concepts may be applied in a project oriented environment for both the private and public sectors. It identifies the project environment success roadmap that consists of four levels leading to project corporate success. Processes that control the dataflow for generating the projects data warehouse are identified and the projects data warehouse contents are defined. The rest of the chapter shows how data mining may be utilised at each project success level to increase the chances of delivering profitable projects that will have the intended impact on the corporate business strategy. The general conclusion is that there is a need to structure and prioritise information for specific end-user problems and to address a number of organizational issues that may facilitate the application of data mining and knowledge management in a project oriented environment. Finally, the chapter concludes by identifying the issues that need to be addressed by private and public sector organizations so that data mining may be utilised successfully in their decision making process.

Clinical Data Mining in the Age of Evidence-Based Practice

Clinical Data Mining (CDM) is a paradigm of practice-based research that engages practitioners in analyzing and evaluating routinely recorded material to explore, evaluate and reflect on their practice. The rationale for, and benefits of this research methodology are discussed with multiple exemplars from health and human service settings. While CDM was conceived as a quantitative methodology evaluating the process, intervention and outcomes of practice, it can support qualitative studies encouraging reflectiveness. CDM was originally employed as a practice based research (PBR) consultation strategy with practitioners in clinical settings, but the methodology has been increasingly used by doctoral students as a dissertation research strategy either by itself or in combination with other research methods. CDM has gained international recognition by both social workers and allied health professionals. The authors present CDM as a knowledge-generating paradigm contributing to “evidence-informed” practice rather than “evidence based practice.”

Data Mining In the Context of Business Network Research

In the field of information technology (IT) enabled business networks and research the traditional data mining approach is theoretically and practically inadequate for knowledge eliction and management requirements in inter-organizational collaborative business environments. The issues are mostly related to fundamentally and philosophically narrow conceptions of the meaning of information, and are grounded to the metatheoretical implications of positivistic, nomothetic and objective view of reality that restricts the feasibility of research oriented application based on them. Here a novel research framework for network-wide knowledge discovery is presented that is based on sociologically anti-positivistic, ideographic and subjective view of society construed from social facts. The theoretical framework is further developed here by synthesizing it with and extracting results from existing research models and artefacts originated in analyzing a variety of business networks (for example, a case study concentrating on modeling the IT enabled service provision of local travel industry value chain). The main contribution here is the explication and elaboration of existing and emerging business network research theories and related stakeholder-level practical considerations focusing on topics such as: multidisciplinary research conceptualizations, information asymmetry reduction by benefiting from contract law oriented functional principles, and network-wide knowledge governance approaches.

Preparing for New Competition in the Retail Industry

A business case presents a retail company facing new competitors and consequently preparing a customer retention strategy. The business environment in which the company was operating prior to the arrival of new competitors can be described as a stable market. Bearing in mind the plans and marketing activities of a competitor retail chain and making use of the data mining methods a system is being devised for the purpose of preventing or at least buffering the churn trend. Development of an early warning indicator system based on data mining methods is also being described as a support to the management in early detection of both market opportunities and threats. Research in data mining could also be concentrated on applying existing data mining techniques to find the best solution regarding practical business problems in the public or private sector. Knowledge regarding how some business cases were solved using data mining techniques could contribute in a better understanding of the nature or data mining nature and help solve specific business issues.

Correlation Analysis in Classifiers

This chapter presents a new method to analyze the link between the probabilities produced by a classification model and the variation of its input values. The goal is to increase the predictive probability of a given class by exploring the possible values of the input variables taken independently. The proposed method is presented in a general framework, and then detailed for naive Bayesian classifiers. We also demonstrate the importance of “lever variables”, variables which can conceivably be acted upon to obtain specific results as represented by class probabilities, and consequently can be the target of specific policies. The application of the proposed method to several data sets shows that such an approach can lead to useful indicators.

Data Mining Methods for Crude Oil Market Analysis and Forecast

In this study, two data mining based models are proposed for crude oil price analysis and forecasting, one of which is a hybrid wavelet decomposition and support vector Machine (SVM) model and the other is an OECD petroleum inventory levels based wavelet neural network model (WNN). These models utilize support vector regression (SVR) and artificial neural network (ANN) technique for crude oil prediction and are made comparison with other forecasting models, respectively. Empirical results show that the proposed nonlinear models can improve the performance of oil price forecasting. The findings of this research are useful for private organizations and governmental agencies to take either preventive or corrective actions to reduce the impact of large fluctuation in crude oil markets, and demonstrate that the implications of data mining in public and private sectors and government agencies are promising for analyzing and predicting on the basis of data.

Data Mining Using Fuzzy Decision Trees

The chapter exposits the strategies employed by the public long-term care systems operated by each U.S. state government. The central technique employed in this investigation is fuzzy decision trees (FDTs), producing a rule-based classification system using the well known soft computing methodology of fuzzy set theory. It is a timely exposition, with the employment of set-theoretic approaches to organizational configurations, including the fuzzy set representation, starting to be discussed. The survey details considered, asked respondents to assign each state system to one of the three ‘orientations to innovation’ contained within Miles and Snows’ (1978) classic typology of organizational strategies. The instigated aggregation of the experts’ opinions adheres to the fact that each long-term care system, like all organizations, is “likely to be part prospector, part defender, and part reactor, reflecting the complexity of organizational strategy”. The use of FDTs in the considered organization research problem is pertinent since the linguistic based fuzzy decision rules constructed, open up the ability to understand the relationship between a state’s attributes and their predicted position in a general strategy domain - the essence of data mining.

Productivity Analysis of Public Services

Productivity is a key success factor in any organization. In order to improve productivity, it is necessary to understand how various factors affect it. The previous research has mainly focused on productivity analysis at macro level (e.g. nations) or in private companies. Instead, there is a lack of knowledge about productivity drivers in public service organizations. This study aims to scrutinize the role of various operational (micro level) factors in improving public service productivity. In particular, this study focuses on child day care services. First, the drivers of productivity are identified in light of the existing literature and of the results of workshop discussions. Second, the drivers most conducive to high productivity and the specific driver combinations associated with high productivity are defined by applying methods of data mining. The empirical data includes information on 239 day care centers of the City of Helsinki, Finland. According to the data mining results, the factors most conducive to high productivity are the following: proper use of employee resources, efficient utilization of premises, high employee competence, large size of day care centers, and customers with little need for additional support.

The Use of Data Mining for Assessing Performance of Administrative Services

The aim of this research was to study the performance of 58 Slovenian administrative districts (state government offices at local level), to identify the factors that affect the performance, and how these effects interact. The main idea was to analyze the available statistical data relevant to the performance of the administrative districts with machine learning tools for data mining, and to extract from available data clear relations between various parameters of administrative districts and their performance. The authors introduced the concept of basic unit of administrative service, which enables the measurement of an administrative district’s performance. The main data mining tool used in this study was the method of regression tree induction. This method can handle numeric and discrete data, and has the benefit of providing clear insight into the relations between the parameters in the system, thereby facilitating the interpretation of the results of data mining. The authors investigated various relations between the parameters in their domain, for example, how the performance of an administrative district depends on the trends in the number of applications, employees’ level of professional qualification, etc. In the chapter, they report on a variety of (occasionally surprising) findings extracted from the data, and discuss how these findings can be used to improve decisions in managing administrative districts.