Maison des Sciences Humaines
11, Porte des Sciences
L-4366 Esch-sur-Alzette / Belval
LISER Salle de Conference, 1st Floor
seminars@liser.lu
Abstract
Large Language Models (LLMs) like OpenAI's GPT series are redefining the boundaries of research, data analysis, and innovation across various sectors. These models can significantly improve research processes, assist in data analysis, and generate insights that benefit multiple fields. Our presentation will introduce some of the capabilities of OpenAI's LLMs, focusing on their practical applications for researchers through an easy-to-follow guide on utilizing the OpenAI API. We will cover three different applications: (1) use basic GPT models to generate data for a skill-classification task, (2) how to fine-tune a GPT model and automate the assigning of multiple potential ISCO codes to job titles, and (3) how to use a state-of-the-art GPT image model for detailed image content description, which can then be used for text analysis later on. Each example illustrates the transformative power of LLMs in solving complex problems and automizing very time-consuming tasks.