Clustering Chemical Data Bases Through Projection of MOS Similarity Measures on Multidimensional Spaces