Abstract: This paper constructs a risk assessment system for unsafe behavior of enterprise personnel by delving into the theory of human factor analysis and classification systems. It introduces the Five-element Connection Number based on set pair theory into the risk assessment model for unsafe personnel behaviors, enabling quantitative evaluation. Furthermore, the paper deeply explores the key factors influencing personnel behaviors in system security. Meanwhile, it analyzes the dynamic development trend of enterprise personnel's unsafe behavior risk using partial connection numbers at each order. Empirical validation demonstrates that this assessment method not only enables quantitative analysis of risk-influencing factors, but also provides reasonably reliable predictions of personnel-induced system risk trends. This offers a scientific basis for decision-making in better understanding and addressing system risks.

Key words: unsafe behavior, set pair theory, Five-element Connection Number, risk assessment, safety management