25th EANN 2024, 27 - 30 June 2024, Corfu, Greece

Comparative Analysis of Large Language Models in Structured Information Extraction from Job Postings

Kyriaki Sioziou, Panagiotis Zervas, Kostas Giotopoulos, Giannis Tzimas

Abstract:

  The recent progress in Large Language Models has opened up new possibilities for their application in different domains. This study focuses on exploring the potential of LLMs in structured information extraction, specifically in the context of job postings. We compare commercial and open-source LLMs to see how well they can extract key information from job postings in Greece's tourism sector. Our goal is to understand the performance differences between these models and assess their general applicability in real-world information extraction tasks. We aim to evaluate and compare the capability of these models in accurately identifying and extracting specific data points such as Job Title, Company, Industry, Location, Soft Skills, and Hard Skills. This research contributes to our understanding of how practical LLMs are in real-world information extraction tasks and highlights the differences in performance among various state-of-the-art models.  

*** Title, author list and abstract as seen in the Camera-Ready version of the paper that was provided to Conference Committee. Small changes that may have occurred during processing by Springer may not appear in this window.