This paper presents a steam energy modeling for paper-making, focusing on factory energy management in the paper drying process and analysis of the response characteristics of the process. A heuristic methodology is developed to model steam energy consumption by efficiently and robustly solving correlations among significant factors. Using paper production and operation data, independent variables were selected to identify the relationships among influencing factors. Data on paper breakage and branching were extracted from facility operation data and analyzed using Spearman correlation. Results of this study showed that the type of paper-making, the cylinder of the drying process, and the pressure and temperature of the steam energy supply were identified as the main influencing factors. Feature factors were extracted, classified into support vector machine (SVM) models, and analyzed using t-distributed stochastic neighbor embedding (T-SNE) dimensionality reduction. The proposed clustering modeling methodology can solve specific problems by extracting additional generalized factors and improving energy efficiency. Based on the analysis of actual process operation data, it is found that controlling steam pressure applied to the cylinder is a significant variable that can reduce steam energy consumption while maintaining appropriate quality. Overall, this study provides valuable insights into the energy management of the paper drying process and offers a methodology for modeling steam energy consumption in paper-making that can be applied to other industrial processes.