Event
Last Fridays Talks: Networks & Graphs
Location
Date
Organizer
Last Fridays Talks
Each last-Friday-of-the-month, we are hosting the Last Fridays Talks, where one of our seven Collaboratories will present insights from their current work. Join us for a discussion on results and recent papers, followed by some socializing afterwards for everyone who wish to attend.
Talk 1
Large-Scale Network Embeddings: Leveraging Social Relationships to Predict Future Life Events
Abstract
Recent work has successfully modeled life trajectories as sequences of events, similar to how we model language. However, understanding the role of social relationships in shaping these paths remains an open challenge. In this talk, I will present an approach for studying how social connections influence life outcomes by integrating social networks with individual trajectories. Using nationwide registry data, we investigate methods for representing life events and social relationships in a unified framework, aiming to discover patterns of social influence in human lives.
Speaker
Christian Vestergaard Djurhuus
Bio
Christian Vestergaard Djurhuus is a PhD student at the Copenhagen Center for Social Data Science (SODAS), funded by the Pioneer Centre for Artificial Intelligence. His research uses registry data and graph representation learning to model human life trajectories. He holds degrees in Artificial Intelligence and Data (BSc) and Human-Centered Artificial Intelligence (MSc) from DTU.
Talk 2
Measuring polarization
Abstract
The perception of growing polarization has become a prominent topic in both academic and public conversations. However, the concept of polarization and the evidence for its increase often remain ambiguous, with many claims resting on anecdotal arguments. Further, any increase in polarization is almost exclusively portrayed in a negative light, despite certain forms and degrees of political disagreement being vital for a healthy democracy. In this talk, I will present how tools from computational social science and network analysis can be employed to measure and analyze different types of polarization in online social media and beyond. These methodologies include quantifying issue alignment, examining mass-elite polarization, and identifying mechanisms for depolarisation. This work demonstrates how the operationalization of concepts from political science can generate compelling theoretical challenges within network science and computational sciences more broadly. By bridging these disciplines, this research offers new avenues for understanding polarization and its underlying mechanisms and evolution.
Speaker
Bio
Mikko Kivelä is an assistant professor at the Department of Computer Science at Aalto University, Finland, where he also obtained his doctoral degree. Before Aalto, he was a postdoctoral scholar at the Mathematical Institute at the University of Oxford. In 2022, he became an Academy of Finland Research fellow. He is a network scientist with an academic background somewhere between mathematics, computer science, and physics. He is known for his foundational work on multilayer and temporal networks within network science. He works with data, models, and algorithms as long as there is a potential for real-world applications. Interdisciplinary problems inspire him, and he has worked on a multitude of research areas with a current focus on social networks.