Chemical ionization mass-spectrometry (CIMS) has been widely used in atmospheric chemistry research to detect oxygenated organic compounds. However, accurate quantification of species which do not have generic standards remains a major challenge for CIMS application, and semi-quantitative methods which have been developed to solve this dilemma usually have very large uncertainties. This work introduced a novel classification approach which greatly enhances the accuracy of semi-quantitative methods and effectively reduces their uncertainties. Although the approach was developed in this work specifically for toluene oxidation products, the concept can also be applied to other oxygenated organic compounds and may significantly enhance the quantification ability of CIMS.