This talk explores how politicians employ personalization strategies on Instagram through both text and images. Drawing on a dataset of more than 12,000 posts from German parliamentarians, the study applies computational methods such as image clustering, zero-shot classification, topic modeling, and text classification. The analysis identifies multi-modal frames that illustrate how personalization operates across visual and textual dimensions, offering new insights into the role of social media platforms in shaping contemporary political communication. 

About the speaker

Yuru Li is a PhD candidate at the Center for Media, Communication and Information (ZeMKI) at the University of Bremen. Her research focuses on computational approaches to political communication, visual communication, and social media study. She is currently affiliated with the Lab for Digital Communication and Information Diversity and has experience in applying computer vision and NLP in social media research.