It is well-known that the composition ratio, physical properties, and formulation ratios of mineral oils used as a plasticizer in hot-melt adhesives significantly affect their compatibility with the base polymer and, consequently, the adhesiveness. In this study, we applied a machine learning algorithm, which can identify patterns from large datasets, to investigate the effects of plasticizer compatibility on the peel strength of hot-melt adhesives composed of various mineral oils. The use of principal component analysis made it possible to visualize the overall trend of large number of viscoelasticity data of hot-melt adhesives. Furthermore, it was confirmed that only varying the composition or physical properties of the mineral oil as used plasticizer to obtain a hot-melt adhesives with a wide range of viscoelastic properties. The relationship between the peel strength and the viscoelasticity of hot-melt adhesives was modeled using machine learning algorithm, it was found that the prediction of the peel strength from only the viscoelasticity information of hot-melt adhesives with a high precision coefficient of determination（r2）of 0.881.