Author(s)

Sinan Jin ¹, Siyuan Zhong ¹, Songhaomin Lin ¹, Qikai Wang ¹

Affiliation(s)

¹ School of Labor Economics, China University of Labor Relations, Beijing, China

Corresponding Author

Sinan Jin

ABSTRACT

As the scale and dimensionality of social survey data continue to increase, research on employee job satisfaction is gradually shifting from traditional statistical analysis to data-driven intelligent modeling. However, labor relations survey data typically includes categorical variables, continuous variables, and various organizational environmental factors, with strong nonlinear correlations between variables, making it difficult for traditional models to fully express these complex relationships. This paper addresses the task of predicting employee job satisfaction levels by constructing a deep learning method that integrates feature tokenization, an FT-Transformer encoder, and an interpretable analysis module. This method transforms survey variables such as union participation, labor protection, enterprise characteristics, individual attributes, and career development into unified feature tokens and utilizes a self-attention mechanism to uncover interaction patterns between different variables. For model evaluation, this paper sets multiple traditional statistical models, machine learning models, and tabular deep learning models as baselines, comparing them in terms of prediction accuracy, class balance recognition ability, and model consistency. The results show that the FT-Transformer-based modeling approach can better adapt to the complex feature structure of labor relations survey data and provides visual support for understanding the roles of union participation, labor protection, and organizational environmental factors in predicting employee satisfaction.

KEYWORDS

Job Satisfaction; Union Participation; Labor Relations; Employee Welfare; Organizational Support

CITE THIS PAPER

Sinan Jin, Siyuan Zhong, Songhaomin Lin, Qikai Wang. Data-driven Modeling for Employee Job Satisfaction Prediction and Union Participation Mechanism Analysis. Social Security and Administration Management (2026). Vol. 7, No. 1, 54-69. DOI: http://dx.doi.org/10.23977/socsam.2026.070108.

REFERENCES

[1] Meybodi A R. Identifying the dimensions of employee experience according to the effect of satisfaction, work place and organizational culture[J]. International Journal of Human Capital in Urban Management, 2024, 9(1): 85-100.
[2] Hemsworth D. Exploring the theory of employee planned behavior: job satisfaction as a key to organizational performance[J]. Psychological reports, 2026, 129(2): 1584-1615.
[3] Loo S H. Key factors affecting employee job satisfaction in Malaysian manufacturing firms post COVID-19 pandemic: a Delphi study[J]. Cogent Business & Management, 2024, 11(1): 2380809.
[4] Hoxha G. Sustainable healthcare quality and job satisfaction through organizational culture: Approaches and outcomes[J]. Sustainability, 2024, 16(9): 3603.
[5] Mengale M S, Rajuskar C, Badarke M V. Bridging the gap: Why industrial workers need unions now more than ever[J]. International Journal for Multidisciplinary Research, 2024, 6(3): 1-5.
[6] Però D, Downey J. Advancing workers’ rights in the gig economy through discursive power: The communicative strategies of indie unions[J]. Work, Employment and Society, 2024, 38(1): 140-160.
[7] Pienaar J. How do job demands and job resources relate to well-being, turnover intention and performance in retail? Insights from Swedish trade union members[J]. The International Review of Retail, Distribution and Consumer Research, 2025: 1-25.
[8] Knox A. Understanding the relationship between job quality and general health in hospitality: Evidence from Australia[J]. International Journal of Hospitality Management, 2026, 137: 104690.
[9] Afifi F. Transformer-based tokenization for IoT traffic classification across diverse network environments[J]. PeerJ Computer Science, 2025, 11: e3126.
[10] Zhou B, Chen W. Enhancing FT-Transformer with a Matérn-driven Kolmogorov-Arnold Feature Tokenizer for Tabular Data-Based In-Bed Posture Classification[J]. IEEE Access, 2025.
[11] Zhu E. Adaptive tokenization transformer: enhancing irregularly sampled multivariate time series analysis[J]. IEEE internet of things journal, 2025.
[12] Li X Y. Industrial data imputation based on multiscale spatiotemporal information embedding with asymmetrical transformer[J]. IEEE Transactions on Neural Networks and Learning Systems, 2025, 36(8): 14937-14948.
[13] LI Y. Integrating evolutionary algorithms and enhanced-YOLOv8+ for comprehensive apple ripeness prediction[J]. Scientific Reports, 2025, 15: 7307. DOI: 10.1038/s41598-025-91939-4.
[14] BAI X, LI Y. DS-ARO: a multi-strategy improved artificial rabbits optimization algorithm for global optimization and corporate bankruptcy prediction[J]. Scientific Reports, 2026. DOI: 10.1038/s41598-026-57561-8.
[15] CHEUNG Y, GUO Z, LI Y. Foliated generative adversarial networks for rolling bearings fault diagnosis[J]. IEEE Transactions on Instrumentation and Measurement, 2026, 75: 1-24. DOI: 10.1109/TIM.2026.3684644.

Sponsors, Associates, and Links

---

• Information Systems and Economics
  
  
• Accounting, Auditing and Finance
  
  
• Industrial Engineering and Innovation Management
  
  
• Tourism Management and Technology Economy
  
  
• Journal of Computational and Financial Econometrics
  
  
• Financial Engineering and Risk Management
  
  
• Accounting and Corporate Management
  
  
• Population, Resources & Environmental Economics
  
  
• Statistics & Quantitative Economics
  
  
• Agricultural & Forestry Economics and Management
  
  
• Social Medicine and Health Management
  
  
• Land Resource Management
  
  
• Information, Library and Archival Science
  
  
• Journal of Human Resource Development
  
  
• Manufacturing and Service Operations Management
  
  
• Operational Research and Cybernetics