Search

We commented on the emergence of knowledge science, including epistemology, knowledge engineering, management science with knowledge management, sociological (soft) systems science, technology management, and technological (hard) systems science. As opposed to philosophic macro-theories of knowledge creation on long historical scale, many new micro-theories of knowledge creation for today and tomorrow emerged since 1990. All such micro-theories take into account the interplay of intuitive and emotional, tacit aspects of knowledge with rational and explicit aspects. A contemporary theory of intuition explains rationally why intuitive and emotive aspects are much stronger sources of creative ideas. There is a qualitative difference between group-oriented organizational processes of knowledge creation in industrial and market organizations and individual-oriented academic processes of knowledge creation; the latter can be described by a Triple Helix of academic knowledge creation. Combining both organizational and academic processes of knowledge creation is the prescriptive Nanatsudaki model of seven creative processes. The importance of diverse elements of these models was empirically supported by the results of a survey of creativity conditions in a Japanese university, using multiple criteria decision making for knowledge acquisition from large data bases.

Artificial neural networks (ANNs) have been widely applied to finance and economic forecasting as a powerful modeling technique. By reviewing the related literature, we discuss the input variables, type of neural network models, performance comparisons for the prediction of foreign exchange rates, stock market index and economic growth. Economic fundamentals are important in driving exchange rates, stock market index price and economic growth. Most neural network inputs for exchange rate prediction are univariate, while those for stock market index prices and economic growth predictions are multivariate in most cases. There are mixed comparison results of forecasting performance between neural networks and other models. The reasons may be the difference of data, forecasting horizons, types of neural network models and so on. Prediction performance of neural networks can be improved by being integrated with other technologies. Nonlinear combining forecasting by neural networks also provides encouraging results.

This paper tries to find interaction rules between members and self-reform rules of members in some scientific research laboratories at a graduate school. The candidate of rules are extracted from the personality description sentences about factors: extroversion, agreeableness, conscientiousness, neuroticism and openness in the big five theory of personality psychology. In this paper, interaction and self-reform rules are not described by personalities themselves; instead, they are described by corresponding phenomena: activity, empathy, persistence, autonomy, and analytical ability which appear in the behavior side. Each member’s initial value is determined from the main 5-factor personality investigation to the member, and the target value is given by the member’s self-declaration. The rule selection is carried out by the genetic algorithm, where a sigmoid function is introduced in the renewal algorithm of the value, whereby a difficulty of setting the renewal parameters is avoided. An analysis is added about the obtained rules, and the validity of this approach and the subjects for future study are discussed. Finally, consideration is extended to the effect of environment on the behavior of members to reinforce the set of rules.

The paper starts from a discussion of the concepts of knowledge management versus technology management, and the emergence of knowledge sciences. This is followed be a summary of recent results in the theory of knowledge creation. Most of them concern diverse spirals of creative interplay between rational (explicit) and intuitive or emotional (tacit) aspects of knowledge. Some of them concentrate on organizational (market or purpose-oriented) knowledge creation, other describe academic (research-oriented) knowledge creation. The problem addressed in this paper is how to integrate diverse spirals of knowledge creation into a prescriptive or exemplar model that would help to overcome the differences between organizational (market-oriented) and normal academic knowledge creation. As such prescriptive approach, the JAIST Nanatsudaki Model of knowledge creation is proposed. It consists of seven spirals, known from other studies, but integrated in a sequence resulting from the experience of authors in practical management of research activities. Not all of these spirals have to be fully utilized, depending on a particular application, but all of them relate to some essential aspects of either academic or organizational knowledge creation. The paper presents Nanatsudaki Model in detail with comments on consecutive spirals. The results of a survey of opinions about creativity conditions at JAIST indicate the importance of many spirals constituting the Nanatsudaki Model. Directions of further testing the Nanatsudaki Model are indicated.

Artificial neural networks (ANNs) have been widely applied to finance and economic forecasting as a powerful modeling technique. By reviewing the related literature, we discuss the input variables, type of neural network models, performance comparisons for the prediction of foreign exchange rates, stock market index and economic growth. Economic fundamentals are important in driving exchange rates, stock market index price and economic growth. Most neural network inputs for exchange rate prediction are univariate, while those for stock market index prices and economic growth predictions are multivariate in most cases. There are mixed comparison results of forecasting performance between neural networks and other models. The reasons may be the difference of data, forecasting horizons, types of neural network models and so on. Prediction performance of neural networks can be improved by being integrated with other technologies. Nonlinear combining forecasting by neural networks also provides encouraging results.

This paper presents an approach to service systems development based on the theory of knowledge construction systems, emphasizing service value co-creation between providers and users. As a concrete example of this approach, this paper introduces a social experiment of developing a community health care system, which requires not only information management but also knowledge management for the success.

This paper proposes a technique to deal with fuzziness in subjective evaluation data, and applies it to principal component analysis and correspondence analysis. In the existing method, or techniques developed directly from it, fuzzy sets are defined from some standpoint on a data space, and the fuzzy parameters of the statistical model are identified with linear programming or the method of least squares. In this paper, we try to map the variation in evaluation data into the parameter space while preserving information as much as possible, and thereby define fuzzy sets in the parameter space. Clearly, it is possible to use the obtained fuzzy model to derive things like the principal component scores from the extension principle. However, with a fuzzy model which uses the extension principle, the possibility distribution spreads out as the explanatory variable values increase. This does not necessarily make sense for subjective evaluations, such as a 5-level evaluation, for instance. Instead of doing so, we propose a method for explicitly expressing the vagueness of evaluation, using certain quantities related to the eigenvalues of a matrix which specifies the fuzzy parameter spread. As a numerical example, we present an analysis of subjective evaluation data on local environments.

This paper presents an agent-based approach to the identification of prediction models for continuous values from multi-dimensional data, both numerical and categorical. A simple description of the approach is: a number of agents are sent to the data space to be investigated; at the micro-level, every agent tries to build a local linear model with multi-linear regression by competing with others, and then at the macro-level all surviving agents build the global model by introducing membership functions. Three tests were carried out and the performance of the approach was compared with a neural network. The results of the three tests show that the agent-based approach can have good performance for some data sets. The approach complements rather than competes with existing Soft Computing methods.

Motorola Inc. first introduced the concept of a roadmap in the 1970s as a kind of strategic planning tool. This paper argues that roadmapping can also be a useful way to manage knowledge in academia and to support scientific research. By applying the principles of interactive methodology, this paper puts forward a new methodology for making personal academic research roadmaps as a way of knowledge management in academia. This methodology is composed of six phases: forming groups, explanation from knowledge coordinators, description of present situation, current status and idealized design reports from every member, research and study schedules, and implementation and control. After introducing the approach, a case study describing the application of the methodology at the Japan Advanced Institute of Science and Technology is given.

The School of Knowledge Science at Japan Advanced Institute of Science and Technology is the first school established in the world to make knowledge creation the core of its scientific research. At this graduate school, knowledge management research is focusing on knowledge conversion theories, knowledge-systematizing models and methods for the development of creativity. It aims at enabling researchers to achieve robust innovation in research frontiers of natural sciences. For this purpose, it is necessary to establish a ‘Ba’ that supports the development and practice of scientific knowledge creation. This paper considers the advantages and disadvantages deriving from the vagueness, depth, diversity and flexibility of the definition of ‘Ba’, stresses the need of redesigning ‘knowledge creation Ba’ using systems concepts, incorporates insights from the structure–agency–action sociological paradigm, and proposes a systems methodology to design and evaluate ‘Ba’ for technology creation in academia, with illustrating evidence from a preliminary survey. Copyright © 2006 John Wiley & Sons, Ltd.

In this paper, we propose a method for detection of local system structures in a complex database. The complex database is viewed as consisting of mixed numeric and nominal attributes, and the local system structure as expressed by “if–then” rules. The detection of local system structures is an important task, and is concerned with inter-dependent issues. The issues involved in the detection of “if–then” rules include finding the objects that share common interests and then finding if–then rules that characterize those objects. To deal with these issues, an agent-based approach is proposed. Each agent has the role of collecting data points (objects) based on their similarity, for mixed data and detecting a rule. The similarity is introduced so that the agent can handle a mixed database. Each agent will occupy a part of the database as its territory according to the predefined algorithm with which agents try to expand or reduce their territories.

This paper describes a multi-agent model built to simulate the process of technological innovation, based on the widely accepted theory that technological innovation can be seen as an evolutionary process. The actors in the simulation include producers and a large number of consumers. Every producer will produce several types of products at each step. Each product is composed of several design parameters and several performance parameters (fitness components). Kauffman’s famous NK model is used to deal with the mapping from a design parameter space (DPS) to a performance parameter space (PPS). In addition to the constructional selection, which can be illustrated by the NK model, we added environmental selection into the simulation and explored technological innovation as the result of the interaction between these two kinds of selection.

Support vector machine (SVM) is a very specific type of learning algorithms characterized by the capacity control of the decision function, the use of the kernel functions and the sparsity of the solution. In this paper, we investigate the predictability of financial movement direction with SVM by forecasting the weekly movement direction of NIKKEI 225 index. To evaluate the forecasting ability of SVM, we compare its performance with those of Linear Discriminant Analysis, Quadratic Discriminant Analysis and Elman Backpropagation Neural Networks. The experiment results show that SVM outperforms the other classification methods. Further, we propose a combining model by integrating SVM with the other classification methods. The combining model performs best among all the forecasting methods.

Recently, an attempt of integration between the theories of fuzzy sets and rough sets has resulted in providing a roughness measure for fuzzy sets [M. Banerjee, S.K. Pal, Roughness of a fuzzy set, Information Sciences 93 (1996) 235-246]. Essentially, Banerjee and Pal's roughness measure depends on parameters that are designed as thresholds of definiteness and possibility in membership of the objects to a fuzzy set. In this paper we first remark that this measure of roughness has several undesirable properties, and then propose a parameter-free roughness measure for fuzzy sets based on the notion of the mass assignment of a fuzzy set. Several interesting properties of this new measure are examined. Furthermore, we also discuss how the proposed approach is used to describe the rough approximation quality of a fuzzy classification.

It is compelling to process scientific literature to support the development of new science and technology. We propose a method to predict new relationships between a starting concept of interest and other concepts by mining scientific literature. In contrast to previous research, we measure the relationship between two concepts not only by their co-occurrence in scientific literature, but also by their sibling relationship in a hierarchical structure of concepts. Therefore, the predicted relationships of concepts obtained with our method are more pertinent to existing relationships within current scientific literature. By introducing a parent set, we propose a measure to evaluate the closeness of two concepts in a hierarchical structure of concepts. In order to deal with the combinatorial problems, we present two ways to limit the number of new relationships, which can be interactively enforced by the user. As in most of the previous research on literature-based discoveries, we choose biomedicine as the field in which to demonstrate our method. A comparison with related research shows that our method exhibits better performance, except in term of Recall. The new relationships predicted by this method can serve as candidates for new research themes, as impetus for inspiration, or as hypotheses to be tested in future.

This paper proposes a multiexpert decision-making (MEDM) method with linguistic assessments, making use of the notion of random preferences and a so-called satisfactory principle. It is well known that decision-making problems that manage preferences from different experts follow a common resolution scheme composed of two phases: an aggregation phase that combines the individual preferences to obtain a collective preference value for each alternative; and an exploitation phase that orders the collective preferences according to a given criterion, to select the best alternative/s. For our method, instead of using an aggregation operator to obtain a collective preference value, a random preference is defined for each alternative in the aggregation phase. Then, based on a satisfactory principle defined in this paper, that says that it is perfectly satisfactory to select an alternative as the best if its performance is as at least "good" as all the others under the same evaluation scheme, we propose a linguistic choice function to establish a rank ordering among the alternatives. Moreover, we also discuss how this linguistic decision rule can be applied to the MEDM problem in multigranular linguistic contexts. Two application examples taken from the literature are used to illuminate the proposed techniques.

In this paper we develop a rule-based model for evaluation of regional environment based on both hard and soft data, where by hard data we mean the statistical measurements while by soft data we mean subjective appreciation of human beings of environmental issues. As people's feeling strongly depends on the social and economical characteristics of administrative regions where they live, we firstly use the hard data concerning these characteristics to do clustering in order to obtain clusters corresponding to regions with the homogeneous social and economical characteristics relatively. We then use the soft data, with the help of data-mining techniques, to develop rule-based models which show association between evaluated items of residents in the clusters. Finally, a relationship between hard data and soft data through an integrated model will be explored. It is shown that the soft data are rather reliable and we should integrate subjective knowledge learnt from soft data into modelling of environmental issues. c2004 Elsevier Science Inc. All rights reserved.

In this paper, we propose a fuzzy-set-theoretic based extension to the landscape theory of aggregation introduced by Axelrod and Bennett in 1993. The significance of landscape theory is that it can provide a deeper understanding of a wide variety of important aggregation processes in politics, economics, and society. To illustrate efficiency of the proposal, we make a simulation with the proposed framework for the international alignment of the Second World War in Europe. It is shown that the obtained results are essentially comparable to those given by the original theory. Consequently, the fuzzy-set-theoretic based extension of landscape theory can allow us to analyse a wide variety of aggregation processes that have previously been considered in a more flexible manner.

In this paper we present interesting relationships between the context model, modal logic and fuzzy concept analysis. It has been shown that the context model proposed by Gebhardt and Kruse [Int. J. Approx. Reason. 9 (1993) 283] can be semantically extended and considered as a data model for fuzzy concept analysis within the framework of the meta-theory developed by Resconi el al. in 1990s. Consequently, the context model provides a practical framework for constructing membership functions of fuzzy concepts and gives the basis for a theoretical justification of suitably use of well-known t-norm based connectives such as min-max and product-sum rules in applications. Furthermore, an interpretation of mass assignments of fuzzy concepts within the context model is also established.

Forecasting exchange rates is an important financial problem that is receiving increasing attention especially because of its difficulty and practical applications. Artificial neural networks (ANNs) have been widely used as a promising alternative approach for a forecasting task because of several distinguished features. Research efforts on ANNs for forecasting exchange rates are considerable. In this paper, we attempt to provide a survey of research in this area. Several design factors significantly impact the accuracy of neural network forecasts. These factors include the selection of input variables, preparing data, and network architecture. There is no consensus about the factors. In different cases, various decisions have their own effectiveness. We also describe the integration of ANNs with other methods and report the comparison between performances of ANNs and those of other forecasting methods, and finding mixed results. Finally, the future research directions in this area are discussed.

This paper presents an agent-based approach to identification of prediction models in two-dimensional data spaces. A number of agents are sent to the two-dimensional data space that people want to investigate. At the micro-level, every agent tries to build a local linear model by competing with others, and then at the macro-level all surviving agents build the global model by cooperating with each other. And a genetic algorithm is introduced for improving the global model built by the agents. Two examples that apply this approach are given. The advantages of this approach are it does not need people to give a certain formula in advance; and most of time, it can give more precise prediction models than those given by traditional methods.

This paper first introduces a new discipline knowledge science and the role of systems science in its development. Then, after the discussion on current trend in systems science, the paper proposes a new systems methodology for knowledge management and creation. Finally, the paper discusses mathematical modeling techniques to represent and manage human knowledge that is essentially vague and context-dependent.

This paper is the first to introduce critical systems thinking into the present study of knowledge and its management. Our preliminary study showed two things about critical systems thinking. First, it provides comprehensive and creative insight into tackling the increasing complexity of human knowledge and knowledge management processes through the organic connection and division of knowledge systems. Second, it encourages the critical use of available methodologies in a coherent way to cope with corresponding subsystems or processes, which breaks down the complexity into convenient units. Based on critical systems thinking, we divided knowledge into two aspects: static substance knowledge and dynamic process knowledge. This division not only provides a concise theoretical framework but also allows knowledge managers and workers to clearly understand the gravity of their work and to selectively utilize well-established methodologies in the practice of knowledge-related activities. This paper serves as an introduction to an application of critical system thinking and total systems intervention in diversified human knowledge topics. Copyright © 2002 John Wiley & Sons, Ltd.

This paper proposes a model for the parametric representation of linguistic hedges in Zadeh’s fuzzy logic. In this model each linguistic truth-value, which is generated from a primary term of the linguistic truth variable, is identified by a real number r depending on the primary term. It is shown that the model yields a method of efficiently computing linguistic truth expressions accompanied with a rich algebraic structure of the linguistic truth domain, namely De Morgan algebra. Also, a fuzzy logic based on the parametric representation of linguistic truth-values is introduced.

KM is increasingly imperative as it is regarded as the key determinant of a firm, industry or country for survival and growth in knowledge era. Varieties of disciplines have made contributions to knowledge and knowledge management. Research focuses on one or more specific fields, but to understand which levels of knowledge processes knowledge management should concentrate on, should be more fundamental than advocacy of knowledge management. Knowledge‐related matters were examined from the viewpoint of systems science. Using critical systems thinking, soft systems thinking etc., a new systematic perspective on knowledge was proposed, aiming to provide a new way of thinking and a useful toolbox on different levels and phases of knowledge management for practical knowledge users.

This paper aims to find a new evaluation method for collecting Kansei and Context data, which is based on a partial comparison process; and a specification method based on customer's target, which is suitable for the special Kansei and Context data obtained from partial comparison process. For collecting Kansei and Context data, we randomly select 5 objects from all objects, and ask people to compare them on each attribute. After many times comparisons, many comparison lists will be obtained. With these lists, we map them into a directed graphic, and with using some graphic processing techniques, we combine all the comparison lists into a whole list without any contradictions, and we map the whole list into a certain range as our evaluated data. To access these special Kansei and Context data, we also discussed two specification methods based on semantic differential method. To test the new method on collecting Kansei data and the specification method, a comparison system and a recommendation system are developed.

Evaluation of services plays an important role in the pre-negotiation process for negotiations over services between autonomous agents. Typically, the service evaluation problem can be regarded as multi-attribute decision making, possibly with uncertainty. This chapter first presents an overview on some decision models, which have been used for evaluating service offers in negotiations, and then discusses the issue of handling uncertainty and imprecision in the service evaluation process by a combined approach for evaluation in order to provide decision support in the pre-negotiation process. Finally, the chapter introduces an evidential reasoning-based model for evaluation of services illustrated by a numerical example.

The objects of data processing in this research are adjectives such as innovative, bright, elegant or cheerfixl expressed by people who are just selecting goods to purchase. The subjective data obtained by a product rating usually has a big variance reflecting vagueness, non-linearity or non-additivity of the feelings of people. Such data should be considered as fuzzy data as well as statistical data. In this paper, a method of constructing a fuzzy factor space is suggested. One of the problems here is how to define the second moment of fuzzy data. After defining a fuzzy metric space, a covering problem is formulated to detect representative adjectives that are used with relatively small differences between people. Finally, the problem of identifying a relation between selected adjectives and pre-selected designs is considered by using a multi-channel adaptive bidirectional associative memory.

Most studies of complex systems take the way of comparing a developed artificial world in the computer with the real world and trying to explore the principles of the complex real world. But, this is not enough to understand complex phenomena and make decisions. This paper emphasizes a soft approach that uses both the logic and the educated intuition of people. Our methodology originates in Sawaragi's shinayakana systems approach that is based on Japanese intellectual tradition. This paper introduces our systems methodology together with our trial of applying it to the global environmental problems.

In this paper, simplified ozone models for potential use in integrated assessment are developed from the EMEP ozone model, which is a single-layer Lagrangian trajectory model. The simplification method uses fuzzy rule generation methodology to represent numerous results of the EMEP model as a response surface describing the source–receptor relationships between ozone precursor emissions and daily tropospheric ozone concentrations.

One recent interest in fuzzy clustering is the simultaneous determination of a fuzzy partition of a given dataset and parameters of assumed models having different shapes that explain partitioned datasets. We propose an objective function to detect linear varieties with different dimensionalities. The noise cluster suggested by Dave is introduced. Since this is not all-purpose method, some techniques are suggested using artificial examples to show how to implement clustering successfully.

The global environmental problem is a new field lying between natural science and political science. In this respect, it is required to build a platform between policy makers and researchers to find a reasonable path to sustainable development. The environment framework model is being developed as such a platform. The paper first outlines this model, and then discusses modeling possibility using the computer network.