Abstract: Soft computing was advocated by Prof. Zadeh as a total technology complementary to the advantages and disadvantages of fuzzy theory, neural network models, genetic algorithms, and so on - a wide variety of topics covered at scientific conferences, in books, in papers, etc. In human-centered information systems, human beings play a central role in information processing. Human information processing involves uncertainty, fuzziness, ambiguity, subjectivity, etc., be dealt with well by soft computing. Human-centered information processing systems are important fields of soft computing. This special issue was motivated by the editors' research project at the Tsukuba Advanced Research Alliance (TARA), University of Tsukuba. The title of this issue is thus similar to the TARA project title, Soft Computing and Human Centered Information Systems. This special issue comprehensively covers soft computing, including chaos theory, rough sets, multisets, as well as fuzzy theory, neural network models, and genetic algorithms. Human-centered information systems are also covered extensively, e.g., human imperfect information processing, human evaluation/judgment, optimal allocation problems, vehicle systems, and human intelligent information processing. This issue focuses on eight papers: The first, A Semantic-Ambiguity-Free Relational Model for Handling Imperfect Information, by Nakata, focuses on imperfect information without semantic ambiguity from the standpoint that an extension of relational models causes semantic ambiguity. This paper proposes an extended relational model in the framework of fuzzy sets and the theory of possibility. The paper formulates set and relational operations as extended relational algebra in the proposed model. The paper is applicable to human imperfect information processing. The second paper, Fuzzy Clustering for Detecting Linear Structures with Different Dimensions, by Umayahara et al., proposes a new objective function and an algorithm for detecting clusters with different dimensionalities. The proposed algorithm improves conventional approaches for detecting linear varieties with different dimensionalities. The paper also uses the noise cluster to deal with extraordinary data. The procedures of the proposed algorithm are demonstrated using numerical examples. The algorithm is useful for human evaluation data processing. As shown by Takahara et al., in An Adaptive Tabu Search and Other Metaheuristics for a Class of Optimal Allocation Problems, an adaptive tabu search for a class of optimal allocation problems uses a set of tables for objects as memory elements in which the search region becomes large, and the structure of memory and the search framework are simplified. This is applied to a class of optimal allocation problems in which small and irregular shapes are placed on a large sheet. The method's effectiveness is compared to results obtained by other metaheuristics. This method is useful for optimal allocation problems faced by human beings. The fourth paper, On Dynamic Clustering Models for 3-Way Data, by Sato, deals with 3-way data consisting of objects, attributes, and times using several clustering models. This paper focuses on the models for 3-way data observed by similarities of objects. The paper proposes models showing exact changes over time by fixing clusters during time. The model configuration is based on fuzzy additive clustering models. Models are modified based on data features. Numerical examples demonstrate that the proposed model shows the movements of objects over time. The fifth paper, A Fuzzy Linear Regression Analysis for Fuzzy Input-Output Data Using the Least Squares Method under Linear Constraints and Its Application to Fuzzy Rating Data, by Takemura, applies a fuzzy linear regression model to the analysis of fuzzy rating data. The paper considers a fuzzy linear regression model with fuzzy input data, fuzzy output data, and fuzzy parameters, since human rating data is usually fuzzy. The paper discusses fuzzy linear regression analysis using the least squares method under linear constraints. The present approach is rather heuristic in that it is an extension of the ordinary least squares method for crisp data. Fuzzy linear regression analysis is applied to psychological studies, i.e., the effect of perceived temperature and humidity on unpleasantness and behavioral intention in fashion shopping. This paper deals with human judgment, considering the human being as a human-centered system. The sixth paper, Study on Intelligent Vehicle Control Considering Driver Perception of Driving Environment, by Takahashi et al., discusses an approach of the design of an intelligent vehicle controller supporting driver vehicle use. The approach considers the interaction of the driving environment, vehicle behavior, and driver expectations of vehicle behavior. The paper uses a multiobjective decision-making model as the intelligent vehicle controller and a fuzzy measures and fuzzy integrals model to reflect driver characteristics. The simulation and experimental results show good vehicle control performance. A vehicle does not move without human control. In this sense, the paper deals with human-centered systems as such. The seventh paper, Determinism Measurement in Time Series by Chaotic Approach and Its Applications, by Fujimoto et al., discusses deterministic chaos. The proposed method, trajectory parallel measure (TPM), distinguishes chaos from embedded time series data. This is simpler than conventional methods and examines only the direction of tangential unit vectors of the trajectory in its neighborhood. This is applied to chaotic time series data with random noise. Fast Fourier transform (FFT) analysis is applied to data to verify the effectiveness of the proposed method. Although FFT analysis cannot distinguish the degree of random noise, the proposed TPM clearly distinguishes it. TPM is also applied to the diagnosis of automobile components. TPM detects abnormal acoustic time series data well. TPM is applicable to fault diagnosis of human-centered systems, e.g., vehicles. The final paper, Linguistic Expression Generation Model of Subjective Content in a Picture, by Iwata et al., proposes a model that expresses subjective contents in a picture given objective information on the picture. Objective information is information on object's location, size, direction, etc. Subjective content is emotions of a human object, the relationship between objects, and object behavior obtained from objective information. Human emotions are recognized from facial expressions using neural network models. Fuzzy reasoning is applied to infer the relationship between objects. Case-based reasoning is used to express object behavior. The effectiveness of the present model is verified by experiments. This paper deals with human intelligent information processing, considering the human being as a human-centered system. We thank Drs. T.Fukuda and K.Hirota, editors in chief of the JACI, for accepting our proposals for this special issue and for their ongoing encouragement during editing. Special thanks are due to all referees for their kind cooperation in helping prepare this issue. We also thank Mr.Y.Inoue for his advice on editing.

Abstract: This paper presents semi-supervised agglomerative hierarchical clustering algorithm using clusterwise tolerance based pairwise constraints. In semi-supervised clustering, pairwise constraints, that is, must-link and cannot-link, are frequently used in order to improve clustering properties. From that sense, we will propose another way named clusterwise tolerance based pairwise constraints to handle must-link and cannot-link constraints in L 2 -space. In addition, we will propose semi-supervised agglomerative hierarchical clustering algorithm based on it. We will, moreover, show the effectiveness of the proposed method through numerical examples.

Abstract: This paper provides new clustering algorithms for data with tolerance. Tolerance is understood in a broad sense, e.g., calculation errors and loss of attribute of data. The concept of tolerance is modified by using new concept of tolerance vector. First, the concept is explained and optimization problems of clustering are formulated using the vectors. Second, the problems are solved using Karush-Kuhn-Tucker conditions. Third, the new clustering algorithms are constructed by using the solutions of the problems. Moreover, the effectiveness of proposed algorithms is verified through some numerical examples.

Abstract: In this paper, two types of semi-supervised fuzzy c -means algorithms are proposed. One feature of proposed algorithms is that they are based on an entropyregularized fuzzy c -means clustering algorithm, while conventional algorithms are based on standard fuzzy c -means. Another feature of proposed algorithms is that the membership updating equation can be obtained explicitly with any fuzzifier parameter value, while in conventional methods, the updating equation must be solved by some numerical method or by a numerically complex refinement with almost all fuzzifier parameters. The influence of supervisor-parameter and fuzzifier parameter on clustering results are discussed based on numerical experiments and compared to the conventional method, demonstrating the feasibility of proposed algorithms.