Information Diffusion in Online Social Networks

We collected large amounts of data from the online social network Twitter and reconstructed the underlying interaction network. This thesis aims to develop and implement a for information diffusion, which we plan to apply on misinformation and conspiracy data.
Master

Over the past two decades, online social networks (OSN) have become a part of the daily life of billions of individuals worldwide. Since their operation inherently generates data, OSNs allow the study of human behavior at massive scales. Simultaneously, advancements in high-performance and cloud computing contributed to the availability of affordable, powerful computer hardware, which makes it possible for researchers to analyze large datasets generated in OSNs. Consequently, OSNs have been the subject of many academic papers in the past decade. Among the topics is the study of the distribution of information on a large scale [1].

Information distribution can be modeled by considering individuals as stateful entities frequently exposed to information accumulated and processed to form opinions, which again serve as new input [2]. The connections through which the exchange of information and thus individuals' influence occur are in many cases part of the core functionality of an OSN and are often called friend- or follower relationships. However, these connections are unweighted and thus signify only the presence of interest. They do not reflect how individuals actually interact or influence each other. Therefore, there is a need for obtaining weighted relationships from the communication that occurs between individuals.

We collected large amounts of data from the online social network Twitter and reconstructed the underlying interaction graphs. Now we need to investigate the diffusion of individual messages.

Goal

This thesis aims to develop and implement a for information diffusion, which we plan to apply on misinformation and conspiracy data.

Learning outcome

AI & machine learning

Big Data Analysis

Complex Network Analysis

Qualifications

You should be open-minded and able to work in an international team of researchers from different institutions. You should have machine learning skills or at least be very interested in machine learning. Moreover, you should have the ability to plan your work schedule independently. However, the most important requirement is motivation.

Supervisors

  • Johannes Langguth
  • Daniel Thilo Schroeder
  • Pedro Lind

References

[1] Vosoughi, Soroush, Deb Roy, and Sinan Aral. "The spread of true and false news online." Science 359.6380 (2018): 1146-1151.

[2] Porter, Mason A., and James P. Gleeson. "Dynamical systems on networks." Frontiers in Applied Dynamical Systems: Reviews and Tutorials 4 (2016).

Contact person