Potential and limitations of using LLMs for personalized advertising

The School of EECS is hosting the following PhD Progress Review 1 Confirmation Seminar:

Potential and limitations of using LLMs for personalized advertising

Speaker: Muhammad Elyas Meguellati
Host: A/Prof Gianluca Demartini

Abstract: This progress report investigates the potential of large language models (LLMs) for generating personalized online advertisements tailored to specific personality traits and the implications of such technology for political advertising, combining insights from two studies.

The first explores the use of LLMs to generate personalized online advertisements targeting the openness and neuroticism personality traits in different online environments. Findings indicate that LLM-generated ads can be as effective as human-written ads in terms of user engagement and preferences, highlighting the potential of LLM-generated personalized content to rival traditional advertising methods with the added advantage of scalability. 

The second study introduces an efficient persuasive text detection model that achieved \textit{stat-of-the-art} on Subtask 3 of SemEval 2023 and applies it to a real-world dataset of Facebook political ads from the 2022 Australian Federal election campaign. This analysis sheds light on the subtleties of persuasive content in political advertising and presents a pragmatic approach to detect and analyze such strategies with limited resources.

The findings of this research have implications for both industry and academia, providing industry practitioners with insights into the potential and limitations of using LLMs for personalized advertising, while contributing to the understanding of persuasion in digital advertising for academics and raising important questions about the implications of AI-generated content for society.

Bio: Elyas is a PhD student at CIRES  in The University of Queensland, his primary interests are: Natural Language Processing, Computation Social Science, and Crowdsourcing. He received his BS.c Degree in Computer Science and Mathematics from University of Batna, Algeria. And his MS.c in Applied Computing from The University of Malaya, Malaysia.