Design of Rule Extraction and Optimization Algorithms in Employee Performance Evaluation in Data Mining Environment

In this study, a data-driven assessment framework integrating multi-criteria decision making, association rule mining and fuzzy clustering methods is proposed to address the scientific and objective needs of rule extraction and optimization in employee performance assessment.The TOPSIS model is improved by triangular fuzzy numbers to realize the objective ranking of business performance.The Apriori algorithm is improved to mine the association rules between competency and performance.The empirical results show that Employee 3 is ranked in the excellent grade with 101.32% task completion rate and 0.8323 relative proximity.The questionnaire results of competency quality had a significant impact on appraisal with a confidence level of 84.3%, while technical title and education were not sufficiently correlated with a confidence level of <30%.The fuzzy decision tree model generated 25 classification rules with a confidence level higher than 63.2%.And combined with the work attitude index with a weight of 0.2913 to complete the comprehensive performance assessment, the results show that the overall performance score of the employees in this enterprise is 0.81362, which is a good grade.This study makes the performance appraisal more objective, precise and efficient, and at the same time expands the application scope of data mining technology in enterprise management.