Employee Performance Evaluation Using NLP and LLM
I am collaborating with Professor Wenqi Fan on a project from industry. The goal is to use machine learning algorithms to assess employees’ interview performance, evaluate their promotion potential, and provide employee ratings. I am responsible for data augmentation and selection.
What I did
Data Augmentation: Augment employee interview data using traditional NLP (random swapping, deletion, insertion, etc.) and a custom evolution prompt-based LLM method, guided by WizardLM and DEITA and other papers, utilizing advanced tools such as Distilable and DataDreamer.
Quantitative Scoring: Developed a few-shot and pair-comparing scoring pipeline with LLMs to analyze and quantify employee interview texts for performance and promotion potential evaluation.
What I get
I have researched cutting-edge work on data augmentation, including traditional NLP and LLM methods, as well as on data selection, gaining a deeper understanding of data engineering.I studied a wealth of knowledge related to LLMs, including training, fine-tuning, and the underlying principles.
I extensively used APIs from mainstream LLMs like GPT and Gemini, accumulating a lot of engineering experience. I solved a series of issues related to parallel processing and checkpoints. I used source code of WizardLM and Distilable to build our pipeline. I’m a good data enginner right now~