In recent years, materials informatics (MI), which applies information engineering to materials science, has been attracting attention. MI is a method to promote the efficient development of new materials by utilizing huge amounts of experimental data and simulation results on materials, and using data science technologies such as machine learning and deep learning. In order to advance MI, it is necessary to extract the search range of materials based on multiple search conditions from the large amount of physical property information of the materials, and to explore while visualizing and understanding the range to be experimentally verified. JMP is a powerful tool for visualizing and analyzing this physical property information. We use JMP's database access function to obtain the feature values of materials and intuitively analyze the data with a dashboard including local data filters. In particular, clustering and model fitting are effective methods for extracting useful knowledge from complex data sets. In addition, even researchers who are not familiar with programming such as SQL and Python can consistently carry out the process from data acquisition to analysis due to the easy operability of JMP. In this presentation, we will introduce examples of MI efforts such as visualization of physical property information using JMP and multivariate analysis.
