An employment tracking and career trajectory analysis system for university graduates

Abstract

The employment market of college graduates is "double increase in supply and demand, competitive polarization" and the traditional employment service system has data islanding and insufficient dynamic tracking ability. This paper proposes a closed-loop employment tracking and career trajectory analysis system based on "data collection-trajectory modeling-intelligent recommendation" technology. The system adopts a hierarchical micro-service architecture, integrates federated learning, differential privacy and dynamic time warping (DTW) algorithms, constructs a career development quality index (Q-value), covering 7 dimensions and 28 indicators such as salary growth and job matching, and develops an industry demand forecasting engine to support the dynamic adjustment of colleges and universities. The empirical study shows that the prediction accuracy of SARIMAX (Seasonal Autoregressive Integrated Moving Average with eXogenous regressors) model on the job demand of AI industry is 27.7% higher than that of the benchmark model, and the personalized trajectory recommendation based on DTW improves the average Q value of science and engineering graduates by 7.6 points (Cohen's d=0.81) 36 months after graduation. In addition, the Gini coefficient of Q value of graduates from 2020 to 2025 rose from 0.28 to 0.46, which highlighted the intensification of employment quality differentiation. This system provides a quantitative tool for solving the problem of "skill dislocation" and optimizing education policies, and has dual values of data security and decision support.