Intra-generational social mobility among educational elites remains underexplored, particularly in the domain of music professionals where institutional transitions intersect with regional and reputational hierarchies. This study proposes a hybrid computational framework that integrates resume mining, quadrant-based flow modeling, and deep learning–driven music recommendation systems to analyze the career trajectories of doctoral graduates in music education from the Yangtze River Delta region.A mobility quadrant system is constructed to categorize elite professionals’ flows (upward, downward, parallel) across university reputation and city-tier levels. Empirical results show that 41.0% of initial postdoctoral transitions exhibit parallel mobility, while 67.6% of laterstage transitions toward academic honors (e.g., Changjiang Scholars) reveal downward movement, suggesting structural stagnation in long-term academic progression. To support mobility inference and profile modeling, we further develop a CNN-based hybrid recommendation model, incorporating Mel spectrograms and user preference embeddings, which outperforms traditional collaborative filtering and SVD models across RMSE, precision, recall, and F1-score.
