Researchers have developed STEP (Sequential Trajectory of Employment Prediction), a novel system designed to recommend career paths by analyzing resumes. STEP utilizes a time-decay Gated Recurrent Unit (GRU) cell to model temporal dynamics and Feature-wise Linear Modulation (FiLM) conditioned on educational attainment to predict a user's next job. The system also incorporates ROUTE, a two-stage contrastive procedure for improving occupation representation. Evaluations on four datasets, including an enhanced JobHop dataset, demonstrate that STEP outperforms existing state-of-the-art baselines in next job prediction. AI
IMPACT This research could enhance job recommendation systems and provide new tools for workforce planning and labor market analysis.
RANK_REASON The cluster contains a research paper detailing a new system and methodology.
- Feature-wise Linear Modulation
- gated recurrent unit
- JobHop dataset
- large language models
- Sequential Trajectory of Employment Prediction
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
