The invention relates to a tobacco package workshop quality data model establishment method, which comprises the following steps: receiving original data input by a user, and preprocessing the original data; obtaining a plurality of quality factors influencing the quality problem in the historical tobacco package workshop; matching the preprocessed data with the characteristics of each quality factor, and entering a corresponding quality factor data set; setting a corresponding data mining algorithm according to the number of quality mode factors contained in each set quality factor; dividingthe corresponding quality factor data set into quality mode data sets according to a corresponding data mining algorithm; and determining a quality mode selection rule according to a set quality modeexpression definition and explanation and a set quality data analysis and evaluation result method, and judging to determine whether the quality mode selection rule is used for next calling.