Consequences of Personalized Information
This research project examines how the automatic personalization of online content on platforms like Google could influence societal processes such as segregation and polarization.
Factsheet
- Lead school Business School
- Institute(s) Institute for Applied Data Science & Finance
- Research unit(s) Applied Data Science
- Strategic thematic field Thematic field "Humane Digital Transformation"
- Funding organisation SNSF
- Duration (planned) 01.10.2024 - 31.03.2026
- Project management Prof. Dr. Ulrich Matter
- Head of project Prof. Dr. Ulrich Matter
- Keywords Polarisation, Algorithm, Personalisation, Web platforms
Situation
With the growing importance of the internet as a source of information, the question of how the dissemination of information through online channels impacts our lives is becoming increasingly significant. There are major concerns that personalized filtering mechanisms (based on the collection of personal data and machine learning) on large web platforms are increasingly leading us to consume only content and information that aligns with our existing worldview. While automated personalization of information consumption certainly offers benefits for individual consumers, there are fears that this development may also entail societal costs, leading to social segregation, political polarization, and radicalization. These effects could undermine democratic institutions and processes, thereby threatening political stability. Such potential negative consequences have recently led to calls for government intervention and ethical guidelines for online platforms. This suggests that the personalization of online information dissemination and its effects should be objectively measurable in some way. This is precisely where this research project comes in.
Course of action
Using a new method based on synthetic internet users (so-called 'bots'), this research project aims to conduct several field experiments to better assess the effects of personalized information delivery on online platforms in terms of segregation, polarization, and radicalization. The proposed method requires a sophisticated software infrastructure, which will be further developed as part of this research project. This method is intended to be refined during the SNSF funding period and applied in three specific field experiments.
