Ismail Hakki Toroslu

Prof. at the Middle East Technical University, Ankara, Türkiye

Prof. Toroslu is with the Department of Computer Engineering, Middle East Technical University (METU) since 1993. He has received his B.S. and M.S. degrees in computer engineering from METU, Ankara in 1987 and Bilkent University, Ankara in 1989 respectively. Prof. Toroslu received his PhD from the Department of Electrical Engineering and Computer Science at Northwestern University, IL USA, in 1993. He has also been a visiting professor in the Department of Computer Science at University of Central Florida between 2000 and 2002. His current research interests include data mining, information retrieval, social network analysis, machine learning and intelligent data analysis. Prof. Toroslu has published more than 100 technical papers in variety of areas of computer science. Prof. Toroslu has also received IBM Faculty Award in 2010.

Title: Bridging AI and Personalization: from social media insights to targeted marketing

Yusuf Muchahit Cetinkaya & Ismail Hakki Toroslu

Abstract:

In today’s digital landscape, integrating artificial intelligence (AI) and personalization is vital for understanding and engaging audiences on social media platforms. In this talk, we investigate scalable AI-driven methods to extract meaningful insights from social media content and transform them into actionable strategies for targeted marketing. The talk begins with Twitter account classification, where AI models leverage metadata to differentiate individual and organizational accounts. This classification establishes a foundation for effective user identification and segmentation, enabling personalized engagement. Building upon this, we introduce a framework for personalized paragraph generation aimed at creating compelling landing pages. Using advanced text generation techniques, AI systems produce coherent, relevant, and engaging content tailored to user needs. The study further explores targeted marketing strategies by applying text analysis methods to design and generate personalized landing pages informed by social media insights. This approach bridges the gap between audience understanding and practical marketing applications. To enable deeper content personalization, the study develops techniques for aspect-based sentiment analysis. These methods facilitate scalable and detailed sentiment evaluation, addressing challenges such as echo chambers and bias while improving content relevance and fairness. Additionally, the research proposes a programmable, stance-directed AI architecture that generates human-like, personalized social media content. This framework aligns AI outputs with user preferences and stances, fostering humanized and context-aware communication strategies. Through theoretical advancements and experimental validation, this work demonstrates how AI can bridge the gap between social media insights, user segmentation, and targeted marketing. The findings contribute to the development of human-centered AI systems that prioritize personalization, coherence, and fairness, unlocking new opportunities for businesses and digital platforms.