Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Optimal Transport for Fairness in Social Media
In the context of social network analysis, whether the purpose is spreading news, a product or a specific technology, the network structure plays a key role in the propagation phenomenon. Individuals getting the information, or the product as first (early-adopters) are in a privileged position, and they play a key role in determining who will receive the information at the end of the spreading process. This translates into the fact that selecting the early adopters in order to maximize the outreach is strategical choice. This optimization problem is known in the literature as social influence maximization. Social influence maximization is relevant in understanding how information propagates in a network of individuals. Many algorithms determining the optimal set of early adopters have been proposed. However, most of them are only based on network measurements such as distance centrality or nodes’ degree and do not consider individuals’ information such as gender, ethnicity, geographical location, thus leading to undesired glass ceiling effects. This information however is of great importance whenever one wants to ensure that all categories in the initial dataset are equitably represented in the final outcome. In these cases, social influence maximization must be juxtaposed with the concept of fairness. The scientific literature related to this field witnesses how automated and greedy algorithms amplify inequalities among different demographic groups and what it means to be fair in information diffusion is still an open question. This project is aimed at getting the student familiar with Optimal Transport Theory as a tool to solve timing problems related to fairness in social media.
The project will include the following tasks:
- A literature review from computer engineering, control engineering, and sociology to understand the state of the art of social influence maximization.
- Building the mathematical background in optimal transport
- Mathematical formalization of the concept of fairness for social influence maximization.
- Exploit optimal transport as a tool for designing a new algorithm driven by social influence maximization and fairness. Algorithm simulation and comparison with respect to no strategic early adopters selection.
- Presentation and final report.
Promising results can potentially be turned into a publication.
**Relevant literature:**
- C. Bunne, L. Papaxanthos, A. Krause, and M. Cuturi. Proximal optimal transport modeling of population dynamics. In International Conference on Artificial Intelligence and Statistics, pages 6511–6528. PMLR, 2022.
- Ana-Andrea Stoica, Augustin Chaintreau, Fairness in social influence maximization, Proceedings of the 2019 World Wide Web Conference, San Francisco, CA, USA, May 2019, pp. 569-574. 2021.
- David Kempe, Jon Kleinberg, Eva Tardos, Maximizing the Spread of Influence through a Social Network
- Wei Chen, Yajun Wang, Siyu Yang, Efficient Influence Maximization in Social Networks, Proceedings of Conference on Knowledge Discovery and Data Mining KDD’09, June 28–July 1, 2009, Paris, France.
- Chen Avin, Barbara Keller, Zvi Lotker, Claire Mathieu, David Peleg, and Yvonne-Anne Pignolet. 2015. Homophily and the glass ceiling effect in social networks.Proceedings of the 2015 Conference on Innovations in Theoretical Computer Science.ACM, Rehovot, Israel, pp. 41–50.
- Ana-Andreea Stoica, Christopher Riederer, and Augustin Chaintreau. 2018. Algorithmic Glass Ceiling in Social Networks: The effects of social recommendations on network diversity. In Proceedings of the 2018 World Wide Web Conference on World Wide Web. International World Wide Web Conferences Steering Committee,923–932.
The project will include the following tasks:
- A literature review from computer engineering, control engineering, and sociology to understand the state of the art of social influence maximization.
- Building the mathematical background in optimal transport - Mathematical formalization of the concept of fairness for social influence maximization.
- Exploit optimal transport as a tool for designing a new algorithm driven by social influence maximization and fairness. Algorithm simulation and comparison with respect to no strategic early adopters selection.
- Presentation and final report.
Promising results can potentially be turned into a publication.
**Relevant literature:**
- C. Bunne, L. Papaxanthos, A. Krause, and M. Cuturi. Proximal optimal transport modeling of population dynamics. In International Conference on Artificial Intelligence and Statistics, pages 6511–6528. PMLR, 2022. - Ana-Andrea Stoica, Augustin Chaintreau, Fairness in social influence maximization, Proceedings of the 2019 World Wide Web Conference, San Francisco, CA, USA, May 2019, pp. 569-574. 2021. - David Kempe, Jon Kleinberg, Eva Tardos, Maximizing the Spread of Influence through a Social Network - Wei Chen, Yajun Wang, Siyu Yang, Efficient Influence Maximization in Social Networks, Proceedings of Conference on Knowledge Discovery and Data Mining KDD’09, June 28–July 1, 2009, Paris, France. - Chen Avin, Barbara Keller, Zvi Lotker, Claire Mathieu, David Peleg, and Yvonne-Anne Pignolet. 2015. Homophily and the glass ceiling effect in social networks.Proceedings of the 2015 Conference on Innovations in Theoretical Computer Science.ACM, Rehovot, Israel, pp. 41–50. - Ana-Andreea Stoica, Christopher Riederer, and Augustin Chaintreau. 2018. Algorithmic Glass Ceiling in Social Networks: The effects of social recommendations on network diversity. In Proceedings of the 2018 World Wide Web Conference on World Wide Web. International World Wide Web Conferences Steering Committee,923–932.
Not specified
Giulia De Pasquale (degiulia@ethz.ch), Nicolas Lanzetti (lnicolas@ethz.ch)
Giulia De Pasquale (degiulia@ethz.ch), Nicolas Lanzetti (lnicolas@ethz.ch)