A Text Mining and Ensemble Learning Based Approach for Credit Risk Prediction

• Published in: Tehnički vjesnik Vol. 30; no. 1; pp. 138 - 147
• Main Authors: Mao, Yang, Liu, Shifeng, Gong, Daqing
• Format: Journal Article Paper
• Language: English
• Published: Slavonski Baod Josipa Jurja Strossmayer University of Osijek 2023; Strojarski fakultet u Slavonskom Brodu; Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek; Građevinski i arhitektonski fakultet Osijek; Faculty of Mechanical Engineering in Slavonski Brod, Faculty of Electrical Engineering in Osijek, Faculty of Civil Engineering in Osijek
• Subjects: credit risks; Data mining; Ensemble learning; Innovations; Prediction models; Regression analysis; Regression models; Risk; strategic emerging industries; text mining
• ISSN: 1330-3651; 1848-6339
• DOI: 10.17559/TV-20220623113041

Traditional credit risk prediction models mainly rely on financial data. However, technological innovation is the main driving force for the development of enterprises in strategic emerging industries, which is closely related to enterprise credit risk. In this paper, a novel prediction framework utilizing technological innovation text mining data and ensemble learning is proposed. The empirical data from China listed enterprises in strategic emerging industries were applied to construct prediction models using the classification and regression tree model, the random forest model and extreme gradient boosting model. The results show that the model uses the technological innovation text mining data proven to have significant predict ability, and top management teamꞌs attention to innovation variables offer the best prediction capacities. This work improves the application value of enterprise credit risk prediction models in strategic emerging industries by embedding the mining of technological innovation text information.