A new icon spotting method for designing a user-friendly GUI is described. Here, each icon can represent continuous and discrete vector data which are possibly high-dimensional. An important issue is icon-margin adjustment or uniforming while the relative positioning is maintained. For generating such GUI, multidimensional scaling, kernel principal component analysis (KPCA) and regularization were combined. This method was applied to a set of city locations and a big data set of web-registered job hunter profiles. The former is used to check to see location errors. There were only little mis-allocations. The latter is a set of high dimensional and sparsely discrete-valued big data in the real world. Through these experiments, it was recognized that the presented method, which combines multidimensional scaling, KPCA and the regularization, is applicable to a wide class of jammed big data for generating a user-friendly GUI.