Improving Skill Extraction from Job Postings Using Synthetic Data and Advanced Language Models

dc.contributor.author	Myronenko, Andrii
dc.date.accessioned	2024-08-23T08:52:10Z
dc.date.available	2024-08-23T08:52:10Z
dc.date.issued	2024
dc.identifier.citation	Myronenko Andrii. Improving Skill Extraction from Job Postings Using Synthetic Data and Advanced Language Models. Ukrainian Catholic University, Faculty of Applied Sciences, Department of Computer Sciences. Lviv 2024, 47 p.	uk
dc.identifier.uri	https://er.ucu.edu.ua/handle/1/4672
dc.language.iso	en	uk
dc.subject	Synthetic Data	uk
dc.subject	Advanced Language Models	uk
dc.subject	Job Postings	uk
dc.subject	Improving Skill Extraction	uk
dc.title	Improving Skill Extraction from Job Postings Using Synthetic Data and Advanced Language Models	uk
dc.type	Preprint	uk
dc.status	Публікується вперше	uk
dc.description.abstracten	Skill extraction involves the automated identification and categorization of skills from textual data, such as job postings, and is important for improving human re- source management and job market analysis. In this thesis, we examine existing research on skill extraction from job postings, highlighting the primary challenges and methods used in the field. Following this review, we propose utilizing Large Language Models (LLMs) for skill extraction, specifically formulated as a sequence labeling task. We will evaluate their out-of-the-box performance and explore the potential of using synthetically generated data by these advanced LLMs to improve the performance of smaller, more efficient Domain-Specific Models.	uk