Statistikseminarium: Hawkes Processes on Social and Mass Media: A Case Study of the #BlackLivesMatter Movement in the Summer of 2020
- Date: 27 October 2022, 12:15–13:15
- Location: Ångström Laboratory, room 64119, zoom link 699 250 9213
- Type: Seminar
- Lecturer: Raazesh Sainudiin (Uppsala University)
- Organiser: Matematiska institutionen
- Contact person: Rolf Larsson
Welcome to this seminar held by Raazesh Sainudiin (Uppsala University) with the title "Hawkes Processes on Social and Mass Media: A Case Study of the #BlackLivesMatter Movement in the Summer of 2020".
Joint work with: Alfred Lindström (Dept. of Maths, UU) and Simon Lindgren (Dept. of Sociology, Umeå University)
Abstract: In this work we study the reports in mass media and interactions in social media during the Black Lives Matter protests following the death of George Floyd, using data from the GDELT project and Twitter. For this we implement the self-exciting counting processes known as Hawkes processes to address our main questions relating to how information and content is spread.
For a macro perspective on how mass media interplay with social media, we implement a bivariate Hawkes process and do a Wald test on the bootstrapped data to find that we reject the null hypothesis that Granger causality does not exist between the data on news reports of street protests in mass media from the GDELT project and the data on interactions in Twitter, surrounding tweets that supported the Black Lives Matter movement. We identify such tweets thorough a detailed network analysis of the Twitter data to identify communities of users who share political views.
(this work is under submission)
Support and Software:
The work was partly supported by a VR Grant titled "Algorithms of Resistance" to Lindgren (Sociology@Umeå) and Sainudiin,
by the Wallenberg AI, Autonomous Systems and Software Program funded by Knut and Alice Wallenberg Foundation to Liam
Solus (Math@KTH), Vera Koponen (Math@UU) and Sainudiin, and by Databricks University Alliance & AWS for cloud computing
infrastructures.
Over three years were spent in data engineering software to be able to ingest terabytes of data in an analytics-ready manner with
additional support from Combient Competence Centre for Data Engineering Sciences, Dept. of Mathematics, UU.
Graner, J., Lindström, A., & Sainudiin, R. (2021). Project MEP (Version 1.0) [Computer software].