©The 2024 International Conference on Artificial Life and Robotics (ICAROB2024), J:COM HorutoHall, Oita, Japan, 2024 Extract tactile qualities from time series data Yasuhiro Suzuki Graduate School of Informatics, Nagoya University, Furocho, Chikusa, Nagoya 464-8601, Japan E-mail: ysuzuki@i.nagoya-u.ac.jp https://ysuzuki.info Abstract We proposed the Tactile Score to describe time-varying tactile sensation by the time variation of vertical force. Tactile quality is essential in hardness/softness, roughness, and temperature. Hardness and softness can be extracted from the shape of the Tactile Score. Roughness can be extracted from the pattern of hardness and softness. The arbitrary time series data can be interpreted as a tactile score by considering the time variation of the vertical force, and the hardness and softness are extracted from the time series data interpreted as the tactile score. This method can extract different features from conventional data science methods. Keywords: Data Science, Time series data, Tactile of Data, Tactile Score 1. Introduction Sensitivity differs from person to person. The same greeting of "Good morning" may be perceived as cheerful by some and noisy by others. There is no correct answer to sensitivity, nor can it be generalized. We can generalize if we take the average of many people's sensitivities. Sensitivities that deviate from the average should not be ignored or directed toward sensitivities closer to the average. A general sensitivity search system is a system in which evaluation criteria for content are modelled for each individual through instructional learning, and each user's evaluation criteria model is used for searching. The following algorithms have been used in a sensory search; color histogram A method to extract features of images and videos; impression analysis using the SD method A method to quantify the impression received from contents by assigning degrees to impression words, learning correspondence between impression words and contents, extracting correlation coefficients between contents, and Extracting correlation coefficients, the distance between contents Projecting the quantified impression words and features of contents onto the feature space and measuring the distance between them. 2. Tactile Score Why do we feel "cheer" from the children's "Good morning"? What is the difference between them? The difference is "the way you say it. How we say it can be characterized by the pitch and volume of our voice, but let us look at the volume of our voice. 576