We propose a new method of semi-structured interview, called “Corkboard Image Mapping,” to visualize pluralistic values behind the stories of individuals and to quantitatively analyze them. The proposed procedure was as follows: 1) extracting main items from stories, 2) mapping to corkboard (Corkboard map) and interpreting it, 3) making an image map based on Multi Dimensional Scaling (MDS map) and interpreting it, 4) comparing the Corkboard map with the MDS map. Interpretation of the maps was made by the interview with each participant. The result of experiment using the proposed method indicated that values were extracted from each stories. It was also suggested that interpretation of axes and mapping enables quantitative analysis of value structure and that comparison between a Corkboard map and a MDS map make it possible to quantitatively discriminate individual mapping strategy. This study demonstrated that common base or dissonance of discussion in a society could be explored, using the proposed method and describing the features of value structures behind stories.
In clinical psychology and psychiatry, projective drawings are used to assess individual personality and holistic understanding of behavior. Especially, Koch developed a tree test, called Baum Test, for the purpose of medical practice and personality assessment. Since there is a poor objectivity and a low reliability in interpretations of drawing pictures when the projective drawing techniques are used, we integrated various image processing techniques to interpret the drawing picture of a mental disorder patient. The proposed analysis in this study was as follow: (1) the gray level histogram moment (GLHM) analysis, (2) the spatial gray level dependence method (SGLDM) analysis, (3) the gray level difference method (GLDM) analysis for the drawing picture, and (4)the singular value decomposition (SVD) method that is a factorization of a rectangular real or complex matrix, with many applications in image processing, (5) the Fourier analysis method that can isolate individual components of a compound waveform for the image, and (6)the clinical interpretation of the drawing picture based on the image analysis techniques. These image analysis techniques for projective tree test were utilized to interpret psychological process of a mental disorder.
This paper presents a method to build a hybrid map in on-line for autonomous robot using a Self-Evolving Modular Network (SEEM). The SEEM is the modular network that have a graph architecture. The module and path in the SEEM are generated in self-organizing manner via a learning process. The proposed method is possible to build a hybrid map in online while moving the autonomous robot. Aim of this work is to develop the system for autonomous building of the hybrid map using SEEM. For this purpose, we performed the following two works: (1) Selecting the backbone algorithm of the SEEM for building the hybrid map, (2) System design for building the hybrid map using the SEEM. In (1), a Growing Cell Structure (GCS), a Growing Neural Gas (GNG), and an Evolving Self-Organizing Map (ESOM) were compared by experiments in order to select the backbone algorithm of the SEEM. As the result, it was suggested that the algorithm of the ESOM is appropriate as the generation mechanism of the SEEM. Moreover, in (2), the hybrid map was built using the SEEM based on the ESOM. As a result, it was suggested the proposal method is possible to build the hybrid map from only vision information.
Lip motion features such as lip width and lip length provide important information to identify persons or recognize commands. Lip motion features have advantage for the theft of registration data because the registration data is changeable like passwords as needed. In addition, such technologies employ a common video camera or a webcam so that they can communalize the interface, which then becomes a system that can be used for both personal authentication and command recognition. However, lip motion features have obscurity, so that the increase in the number of users may result in reduced identification and recognition accuracy. Therefore, developing a method based on result of analysis of lip shape feature is important in order to establish a high reliability personal authentication and command recognition system. Lip shape is a unique body feature to individuals. In this paper, we analyze lip shapes in local region, and develop a grouping method for personal authentication and command recognition system. Lip area is divided into local regions A to C. Region A is a rectangular region that subsumed oral fissure. Region B and region C are the rectangular areas over and under the region A, respectively. After dividing regions, we measure the size of six parts in each region and calculate lip shape features. Next, we analyze lip shape features( i.e., upper and lower lip thickness, oral fissure shape, and aspect ratio). The analysis results show each features were divisible into three groups. Therefore, 27 categories were set and classification was done with fuzzy-reasoning. In the experiment for 52 persons, over 80 percent of subjects were well classified as similar categories. Experiment results suggest that grouping method using the proposed features is useful to narrow down authentication targets. In addition, it was shown that the proposed method classified lip shapes more clearly than k-means clustering.
Consumption behaviors of some goods, such as books, compact disks and so on, are affected by others' consumption behaviors( externalities), which are represented by the amount of sales and their ranking positions. Although numerous attempts have been made to the externalities with micro reference groups, in particular buzz marketing, little attention has been paid to those with macro reference groups. This paper proposes a hierarchical bayesian model with the externalities. We estimate the parameters in our model, based on responses to a survey on books, which is used for conjoint analysis. By using the results of estimation, we make a simulation on book sales, taking the externalities into account. We clarify what distributions the externalities are and what attributes are affected by the externalities. In the book market, female has been more affected than male by others' behaviors. In addition, the externalities have more effects on older people. We illustrate that the number of total sales is higher when externality is present than when absent in our model, using numerical examples.
The aims of this paper are to establish the shape space theory which deals with a set of shapes, and to develop a shape space estimation method which does not require the prior knowledge about the topology of shapes. To achieve these aims, first we developed a Self-Organizing Map( SOM) that is free from topological restriction, namely, Topology-Free SOM( TFSOM). Then we built TFSOM into the higher-rank SOM, which is called TFSOM×SOM in this paper. TFSOM×SOM was applied to handwritten digit recognition task, and it showed about 95% accuracy in recognizing digits written by 1000 people, after the TFSOM×SOM was trained by the digits of only 10 people. Since the proposal method is not specialized to line-drawing, it can be applied to wider tasks relevant to shapes.