Search papers, labs, and topics across Lattice.
The paper introduces a two-step approach for next occupation prediction using LLMs, where a reason generator first infers a user's preference based on their history, and then an occupation predictor recommends the next job. To address the misalignment of LLMs with career paths, the authors fine-tune small LLMs using high-quality "oracle reasons" generated by a LLM-as-a-Judge. Experiments demonstrate that this approach enhances prediction accuracy, outperforming unsupervised methods and achieving comparable results to fully supervised methods, while also showing the importance of reason quality.
Fine-tuning a single LLM to both reason about and predict future occupations surprisingly beats using two separate fine-tuned LLMs for each task.
In this work, we develop a novel reasoning approach to enhance the performance of large language models (LLMs) in future occupation prediction. In this approach, a reason generator first derives a ``reason''for a user using his/her past education and career history. The reason summarizes the user's preference and is used as the input of an occupation predictor to recommend the user's next occupation. This two-step occupation prediction approach is, however, non-trivial as LLMs are not aligned with career paths or the unobserved reasons behind each occupation decision. We therefore propose to fine-tune LLMs improving their reasoning and occupation prediction performance. We first derive high-quality oracle reasons, as measured by factuality, coherence and utility criteria, using a LLM-as-a-Judge. These oracle reasons are then used to fine-tune small LLMs to perform reason generation and next occupation prediction. Our extensive experiments show that: (a) our approach effectively enhances LLM's accuracy in next occupation prediction making them comparable to fully supervised methods and outperforming unsupervised methods; (b) a single LLM fine-tuned to perform reason generation and occupation prediction outperforms two LLMs fine-tuned to perform the tasks separately; and (c) the next occupation prediction accuracy depends on the quality of generated reasons. Our code is available at https://github.com/Sarasarahhhhh/job_prediction.
