Swedish Conference on Computational Social Science

Uppsala, 23-25 April 2024

The objective of the conference is to bring together researchers from different disciplines interested in using or in the use of computational methods to study or address societally relevant phenomena. The conference is targeted to researchers affiliated with Swedish institutions, to contribute to the establishment of the Computational Social Science (CSS) community in Sweden.

Topics of interest include, but are not limited to:

  • Computational studies of society (cultural patterns and dynamics, politics, media, …).
  • Computational methods with applications in social science (social network analysis, natural language processing, computer vision, agent-based models, statistics, causal inference, machine learning, …).
  • CSS method development and methodology (validity, replicability, model assessment, statistical theory, …).
  • Digital sources of social data (APIs, population-scale data, corpora, image data, audio data, …).
  • Ethical, legal, and societal aspects of CSS.

The conference will be held in Uppsala on April 23-25, 2024 (lunch to lunch). To have an inclusive event, there will be no registration fee; coffee breaks, lunches, and a social dinner will be provided at no cost. Some funding for travel and accommodation is also available.

Venue: The conference and pre-conference events are located at the University Main Building, at Blåsenhus, at the Ångström laboratory, and at Engelska Parken. Please check the programme below for the details.

Registration is now closed.

For questions, please contact us at: it-uucss-info@lists.uu.se.

Programme

Pre-conference training workshops (23 April 2024)

09:00-11:45 Workshop A: Natural Language Processing and Sentiment Analysis (Location: Centre for Digital Humanities and Social Sciences, Engelska Parken)

09:00-11:45 Workshop B: Analysis of Feature-Rich Networks (Location: Room 101172, Ångström laboratory)

Day 1 (23 April 2024, University Main Building)

12:00-13:00 Lunch (sandwiches) & Registration (outside Room IV)

13:00-13:15 Conference opening (Room IV)

13:15-14:15 CIRCUS keynote: Elisa Omodei, Department of Network and Data Science, Central European University

14:15-14:20 Leg stretcher

14:20-15:05 Lightning talks (L1 8x5min)

15:05-16:00 Coffee break with speed dating

16:00-16:45 Parallel presentation sessions (S1A, S1B)

Day 2 (24 April 2024, Blåsenhus)

09:15-10:15 eSSENCE keynote: Matti Nelimarkka, Centre for Social Data Science, University of Helsinki (Slides)

10:15-10:45 Coffee Break

10:45-11:25 Parallel presentation sessions (S2A, S2B)

11:25-11:30 Leg stretcher

11:30-12:10 Parallel presentation sessions (S3A, S3B)

12:10-13:15 Lunch at Hotel von Kraemer (map)

13:15-14:15 AI4Research keynote: Peter Hedström, Institute for Analytical Sociology (IAS), Linköping University

14:15-14:20 Leg stretcher

14:20-14:50 Lightning talks (L2 6x5min)

14:50-15:30 Coffee Break (Theme: challenges in CSS)

15:30-17:00 Parallel presentation sessions (S4A)

18:30-21:00 CDHU social dinner at Nedre Slotts (map)

Day 3 (25 April 2024, University Main Building)

09:00-09:10 The future of CSS in Sweden: preparation (Room XI)

09:10-10:10 The future of CSS in Sweden: discussion in small groups, with coffee (Café Alma)

10:10-11:00 The future of CSS in Sweden: summary and action plan (Room XI)

11:00-12:00 Networking

12:00-13:00 Lunch at Stationen (map)

Sponsors

The event is supported by:

  • CIRCUS - Centre for Integrated Research on Culture and Society.
  • CDHU - Centre for Digital Humanities and Social Sciences.
  • eSSENCE - The e-Science collaboration.

Keynotes

CIRCUS keynote

Speaker: Elisa Omodei, Department of Network and Data Science, Central European University

Title: Computational methods for the Sustainable Development Goals: the case of food security

Location: Room IV, University main building

Abstract: In a rapidly changing world, facing an increasing number of socioeconomic, health and environmental crises, computational methods can help us quantify vulnerabilities and monitor progress towards achieving the UN Sustainable Development Goals. In this talk, I will first provide a non-exhaustive overview of the main areas of applications where non-traditional data and computational approaches have shown their potential for social impact, and I will then deep-dive more specifically into my work on predicting food insecurity from conflict, weather, and economic data.

eSSENCE keynote

Speaker: Matti Nelimarkka, Centre for Social Data Science, University of Helsinki

Title: Theory, validity and reliability as the new frontier of computational social science? (Slides)

Location: Eva Netzelius Hall, Blåsenhus

Abstract: The field of computational social science is experiencing a period of robust growth: scholarly communities are flourishing, accompanied by a proliferation of scholarly articles and books, and the emergence of intriguing applications of sophisticated methodologies to examine societal phenomena. At the same time, there appears to be a tendency to overlook social theory within computational social science. Nonetheless, social theory distinguishes computational social science from purely computational disciplines such as data science. Its cruciality has been underscored in numerous scholarly works, but still there seems to be less clear idea what is the role of social theory in computational social science. We shall investigate this discussion and open up the question what is theory and how it helps us to do even more awesome work. Similarly, validity and reliability represent cornerstone principles within the realm of social inquiry. Nevertheless, it seems that computational social science sometimes takes shortcuts and for example a discussion between critical algorithm studies - highlighting various issues with these methods - and computational social science - using said methods - seems less common. I will discuss potential implications of such tendencies and prompts reflections on these.

AI4Research keynote

Speaker: Peter Hedström, Institute for Analytical Sociology (IAS), Linköping University

Title: On CSS and the Future of Sociology

Location: Eva Netzelius Hall, Blåsenhus

Abstract: Sociology is a broad social science discipline that embraces a range of different theoretical and methodological traditions. One important common theoretical denominator in much of sociology dating all the way back to the founders of the discipline such as Max Weber, has been the focus on social interactions and their importance for explaining various social phenomena. As noted by Robert Merton and many others, however, there often has been a problematic gap between sociological theory and empirical sociological research. While sociological theories have emphasised the importance of social interactions, empirical research typically has been based on types of data, such as survey data, that typically includes no information on whom the individuals interact with or the broader social networks in which they are embedded. This talk focuses on the extent to which the type of large-scale data and computationally intensive methods of computational social science can help to close the gap between theory and research. By providing relevant analytical tools to rigorously address sociology’s core questions, it is argued that computational social science has the potential to advance sociology in a similar way that the introduction of econometrics advanced economics during the last half century.

Parallel presentation sessions

Session S1A (social processes and social media)

Location: Room VIII, University main building

Chair: Davide Vega

  • Social Influence and the Emergence of the Unexpected: How Novel Music Spreads on Spotify (Martin Arvidsson)
  • Social judgment in online communities: user vocality and disagreement (Diletta Goglia)

Session S1B (methods: visual political communication)

Location: Room IV, University main building

Chair: Alexandra Segerberg

  • Political Visual Communication across Social Media: What are Computer Vision Tools Useful For? (Michael Bossetta)
  • Visual Structural Topic Modeling for the Analysis of Political Communication: A Deep Generative Approach (Matias Piqueras)

Session S2A (social media)

Location: Eva Netzelius Hall, Blåsenhus

Chair: Alexandra Segerberg

  • The re-mediation of legacy and new media on Twitter: A six-language comparison of the European social media discourse on migration (Anamaria Dutceac Segesten)
  • Studying online polarisation during #COP meetings with multilayer networks and topic modelling (Alessio Gandelli)

Session S2B (methods: machine learning)

Location: Room 11:131, Blåsenhus

Chair: Måns Magnusson

  • Transformer-based control variates for efficient finite sample statistics for textual data: public hate crime estimation (Hannes Waldetoft)
  • “Interpretable Configurational Regression” (ICR): A Machine-Learning Method Designed for Social Scientists (Christopher Swader)

Session S3A (word embeddings)

Location: Eva Netzelius Hall, Blåsenhus

Chair: Sara Stymne

  • Increasing numbers, an increasing threat? Using word embeddings to study the relationship between threat-based stereotypes and out-group size (Rojan Karakaya)
  • Word embedding uncertainty estimation (Väinö Yrjänäinen)

Session S3B (knowledge and uncertainty)

Location: Room 11:131, Blåsenhus

Chair: Davide Vega

  • Modelling curiosity and intrinsically motivated knowledge acquisition (Felix Thiel)
  • Social networks with uncertain edges (Xin Shen)

Session S4A (organisations and academia)

Location: Eva Netzelius Hall, Blåsenhus

Chair: Mikael Bask

  • Towards Representation and Reasoning of Organisational Routines (David Johnson)
  • Applying Intersectional MAIHDA to Studying Social Inequalities in Academic Achievement (Anna Soloveva)

Leg stretcher

  • Social and cultural closure in awarding the Nobel Prize in Physics (Jacob Habinek)
  • Skill, Status and the Matthew Effect (Mikael Bask)