基于数据挖掘的高校教师离职倾向预测模型构建研究

中图分类号：G451 文献标识码：A DOI：10.16400/j.cnki.kjdk.2025.28.035

Research on the Construction of Turnover Intention Prediction Model for University Teachers Based on Data Mining

CUI Liang， WANG Chun（Changchun Finance College， Changchun， Jilin ）

AbstractThe phenomenon of colege teachers leaving poses a challenge to the stable development of the higher education system，and accurate prediction of turnover intention has become a key link in humanresource management. Data mining technology can extractresignation risk characteristics from teacher behavior paterns，careedevelopment trajectories，and organizational interactions by integrating heterogeneous data from multiple sources.The construction ofa turnover intention predictionmodel requires four stages： data collection and preprocessng，feature extraction and variablescrening，algorithm fusiondesign，and performance evaluation.It forms acomplete closed-lop system from risk identification to intervention measures. The prediction model plays an importantrole in the stability management of university faculty.It not only improves theaccuracy of university faculty turnover risk identification，but also providesscientificbasis foruniversities tofomulatepersonalizedretentionstrategies，therebypromotingthe transformation of university human resource management towards refinement and intelligence.

Keywordsuniversity teachers; turnover intention; data mining; prediction model

高校教师作为知识创新与传播的核心力量，其队伍稳定性直接关系到高等教育质量与学科建设水平。（剩余4455字）