Masahiro Sawai

Statistical method for classifying cries of baby based on pattern recognition of power spectrum

We develop a novel method applicable to classify the causes of crying infant based on pattern recognition of power spectrum of the cry. In our frame relied on F-value, it is available in power spectrum to order a statistical significance of frequency points which will contribute to the classification of cries. We verify performance of the method by taking the painful cries of infant with the genetic disease (ADEL). The result of our method achieves an excellent prediction. We also give a discussion on the relation between the set of frequency points extracted and the formants of cries.

Relation between impressions of a touch panels' coloration and operation

The relationships between the colors used in the design of touch panels and the operation of electronic devices are discussed, and experiments that investigated the effects of color in the designs of touch panel interfaces on their operation and on the impressions of users have been conducted. Subjects tried mental arithmetic tests on 12 types of screen which had different color images. The “time it took to perform the task,” the “number of correct answers,” the “number of incorrect answers,” and the “number of timeouts” were measured and counted. After the tasks were administered, the subjects answered questionnaires about their impressions of the interfaces. The average number of correct answers was higher when using “cool-casual” than when using the other colorations. Furthermore, there was a positive correlation between the number of correct answers and a preference for the display coloration. There was a strong relationship between the coloration of the interface and the speed and accuracy of the operation. The results of the study lead us to expect that both the usability and the accuracy of operations could be improved by designing the coloration of the touch panel interface appropriately.

Conceptual Cognitive Scheme of Kansei on Referring to the Use of Human Language

While Kansei has a variety of meanings, it is important to integrate those various meanings and form an essential model of Kansei being applicable to engineering. In this paper, we propose a conceptual cognitive scheme of Kansei assuming that the basic notion of Kansei consists of two functions in mind, i.e.,1) how to feel the world and 2) how to represent it. It is a unique point of the scheme to suggest such a relation between Kansei and the use of language that can be applied to analyze Kansei being expressed in human language or to represent Kansei by means of human language. We also give a discussion on the interpretation of the two adjective phrases specified as the candidates for testing Kansei quality of the sounds reproduced by custom-made speaker unit. In conclusion, the proposed two adjective phrases can be properly interpreted based on the art of metaphor.

A Study about Recommendation Based on Underlying Desires

Today, consumers have been able to buy various items with development of e-commerce sites and so on, while it made difficult to find items suitable for the consumers preference from many items. Because of such situation, recommendation systems that can present items suitable for users preference have been required. When a person prefers an item, but hesitates what to do, it is an important criterion whether the persons desires is satisfied by the item or not.Thus, we analyzed relationship between consumers desires and preferred items by Evaluation Grid Method,recommending items based on constructed Evaluation Grids.Moreover we investigated effectiveness of recommendation based on desires in our experiment and discussed about cross-category recommendation such as recommendation of item that have different attributes or categories.

Applicability of F-value to Classification Problems with a Numerous Number of Explanatory Variables: Toward Classification Problems in Biometric and Kansei Engineering

A Feature Extraction algorithm for systems with a numerous number of explanatory variables is developed The algorithm(FSM) achieves perfect performance in classification in gene expressions and hearing problems.