Science

Welcome to the Science page. Here we share our latest research and findings.

React Flow

Scientific Fields

Data Science

Most of what we do at SynoSys is based on data. We share an affection for large scale, heterogeneous datasets obtained in natural experiments, i.e. data that was not generated with a scientific intention but rather as a byproduct of typically digital technologies, like cell phones, individual mobility patterns, online platforms, social media and citizen science projects. In data science we emphasize the explorative nature of science. Metaphorically speaking, we do data sciences without a map - just with a compass, curiosity and intuition.

Citizen Science

SynoSys is not only about bridging gaps between traditionally unconnected scientific disciplines. We are also about removing barriers between the scientific community and the public. Digital technologies of today permit the creation of citizen science projects and citizen science experiments in which the public contributes to large scale data donation projects, engage in the scientific discourse, participate on all levels in the knowledge creating process. A good example is the Corona Data Donation Project, which we launched during the COVID-19 pandemic.

Network Science

Network science pervades our research. It offers new perspectives on many complex systems in biology, social science and technology. Network science can reveal hidden structures in complex systems and, more importantly, point out similiarities between systems that may seem unrelated at first glance and identify universalities and fundamental design principles. Network science plays a key role in our research on human mobility, infectious disease dynamics, and computational social science.

Dynamical Systems Science

Almost all systems we investigate at SynoSys are dynamic. We are interested in complex dynamical systems, e.g. the spread of infections in a population, evolutionary processes, behavior and adaptation, collective behavior, pattern formation, to name a few. Unsurprisingly, dynamical systems science is a core methodological area at its center. Particularly in conjunction with network science we are trying to understand the connection between network structures and their imprint on dynamical that are in turn driven by them.

Machine Learning

Machine learning methods are widely used to categorize and analyze often unstructured data. These approaches enable, for example, dimensionality reduction in high-dimensional observational datasets from wearable devices, pattern recognition in time series data, and the processing of large volumes of text. In particular, natural language processing (NLP) has become a key application area of modern machine learning. NLP techniques, including large language models (LLMs), allow for the systematic analysis of large-scale unstructured text corpora, e.g., from social media, to identify topics, assess sentiment, and examine their temporal dynamics.

Social Sciences

Modern societies are deeply shaped by digital communication networks. We analyze social media dynamics, polarization, and the functioning of democratic processes through the lens of social networks and computational social science. By developing advanced computational methods for social media analysis and exploring the use of AI in the social sciences, we aim to better understand how information flows, opinions form, and collective behavior emerges in connected societies.

Health Sciences

We study health and disease through the lens of complex systems. Our work combines digital epidemiology, infectious disease dynamics, and large-scale health and wearable sensor data to better understand how diseases spread and how health states evolve in populations. A particular focus lies on emerging health challenges such as Long Covid, where integrating diverse data streams and modeling approaches can reveal hidden patterns and inform more effective public health strategies.

Life Sciences

Biological systems are fundamentally collective and networked. We investigate cooperation in biological systems, the role of networks in biological organization, and the principles underlying collective behavior in animals. Our research also explores the structure and function of fungi and the stability of microbiological consortia, aiming to understand how interactions across multiple scales—from molecules to ecosystems—give rise to resilient living systems.

Research Topics

Infectious Disease Dynamics

The dynamics of infectious diseases has been one of our key research interests for the past two decades. During the COVID-19 pandemic we were kept very busy in this domain. We develop large scale models for spreads of infectious diseases, especially on a global scale, and focus on closing the loop between contagion dynamics -> information dynamics -> behavioral responses. In this context, we devote a lot of effort to measuring relevant behavioral responses using digital tools, cell phones, pervasive data, and data obtained in natural experiments.

Democracy and Digital Media

Digital media has changed the way public spheres function around the world, specifically in the information environment of the internet. Moving from mostly one-to-many communication to a many-to-many system has increased its complexity. In addition, big platforms are curating information flows algorithmically and optimize them for maximum engagement. Our goal is to better understand how those transitions are affecting democracies globally and political behavior in particular. Phenomena such as affective polarization, the rise of populism, the spread of misinformation and diminishing trust in institutions are potential developments of concern. For us, data tools, but also causal inference and experimentation, are essential to achieving this goal of empirically describing the complex mechanisms at play between human behavior, technology and politics.

Field Studies on Online Platforms

New data sources make it possible to quantify human behavior, especially on social media. For example, the structures of social networks can be captured with high precision, and over time, content such as texts, but also images, can be categorized and linked to people or social reactions. Only recently, Article 40 of the Digital Services Act mandated access to such data for research in the European Union, and several platforms have already (re)opened their access in response. When such data is linked to individuals’ survey responses, it becomes a powerful tool to describe, for example, the relationship between online behavior and political attitudes. Furthermore, online platforms enable a variety of novel experiments, e.g. through browser add-ons that change parts of the online environment or game-based approaches to study collective behavior. With these methods, we aim to cover the spectrum between experimental control and ecological validity in the study of human behavior online.

Human Mobility

A number of research projects at SynoSys revolve around human mobility and mobility network analysis. We are interested in disentangling hidden structures within mobility networks, how structural features in mobility networks impact dynamical processes that evolve on them, e.g. how structures in the global air-transportation network shape, how pandemics spread across the globe and how mobility network structures respond to disruptions and change over time.

Digital Epidemiology

Digital epidemiology is an active and growing research area at SynoSys. We have designed and implemented several digital projects in the area of epidemiology and public health, ranging from a nationwide mobility monitor to assess behavioral responses to changes in the dynamics of COVID-19 pandemic, the launch of the Corona Data Donation Project in which more than 500.000 participants donated more than two years of data collected on wearables, their fitness trackers and smart watches, to studies on the effects of digital tracing apps in the context of COVID-19. Currently we are focusing on the use of wearable devices to improve our understanding of Long- and Post-COVID.

Cooperative Phenomena

Recently, we started exploring mechanisms of cooperative phenomena in biological and social systems. We are particularly interested in the emergence of cooperative behavior among unlike entities, cooperation as a driver for innovation and selection processes beyond concepts of species and individuals. Using methods from network science and dynamical systems science we investigate entangled evolutionary processes, structural stability of ecological networks and how cooperation may make systems more robust.

Covid-19 & Long-Covid

During COVID-19, our research was dominated by the pandemic. The research ranged from modelling the global spread on the worldwide aviation network and import risk for different countries and regions during the early phase of the pandemic, modelling the impact of non-pharmaceutical interventions, the impact of digital contact tracing, and various digital epidemiology projects. We now focus on understanding phenomena in the aftermath of the pandemic, particularly Long- and Post-Covid Conditions. The foundation of this research is the massive data collected in the Coronal Data Donation Project.

React Flow