These are the webpages of the Deliberation & Argumentation Special Interest Group (DeAr SIG) of the Hybrid Intelligence Center.
For more information or to be included in the group, get in touch with the SIG coordinators: Davide Grossi and/or Enrico Liscio.
When: 4 October 2024, 16:00 CEST [Event will be rescheduled in September]
Speaker: Julia Romberg, Heinrich Heine University Düsseldorf
Title: Review of Computational Text Analysis Methods to Support the Evaluation of Contributions in Public Participation
Abstract: Engaging citizens in decision-making processes is a widely implemented instrument in democracies. Such public participation processes serve the goal of achieving a more informed procedure to potentially improve the process outcome and increase the public acceptance of decisions made. As public officials try to evaluate the often large quantities of citizen input, they regularly face challenges due to restricted resources. For textual contributions, the most common form of citizen input, natural language processing offers the prospect of automatic support for evaluation. In this talk, I will give an overview of computational text analysis methods that have been proposed to support public participation efforts in the literature so far. While some promising approaches exist, there remain important challenges before these can offer reliable support in practice. Against this background, I will discuss a number of avenues that future research should pursue to develop practical solutions.
When: 28 Mar 2024, 14:00 CET
Speaker: Khalid Al Khatib, University of Groningen, the Netherlands
Title: Analyzing Deliberation Strategies in Wikipedia Discussions
Abstract: Wikipedia, the collaborative online encyclopedia, relies on the collective efforts of its users to maintain and enhance the quality of its content. To achieve this goal, Wikipedians engage in complex deliberations on dedicated talk pages, employing various strategies to reach consensus on articles' structure and information accuracy. This talk delves into the fascinating world of these discussions through a series of ongoing and completed studies, demonstrating the use of various attributes such as frames, dialogue acts, and the types of claims and evidence used by editors. By exploring these attributes, we aim to understand human deliberation to improve AI's ability to engage in meaningful discussions and enhance human decision-making processes.
When: 14 Feb 2024, 14:00 CET
Speaker: Enrico Liscio, TU Delft, the Netherlands
Title: Context-Specific Value Inference via Hybrid Intelligence
Abstract: Values, e.g., benevolence and self-determination, are the abstract motivations that explain and justify human behavior and opinions. For policy-makers, crafting policies that align with citizens' value preferences is crucial to long-term effectiveness. However, value preferences differ across individuals and are dependent on context. Techniques must be devised to estimate individual and contextual value preferences of citizens. We refer to this as the value inference challenge, which is the focus of this talk. Value inference entails several challenges and the related work is scattered across different AI subfields. We present a comprehensive overview of the value inference process by breaking it down into three distinct steps (value identification, classification, and estimation), introducing the related challenges and the methods we developed to address them. However, we recognize that relying solely on AI methods to infer values may not yield accurate estimates, due to the implicit nature of human value preferences. Humans must be actively engaged for a successful value inference. To this end, we propose a Hybrid Intelligence (HI) vision where human and artificial intelligence complement each other during the value inference process. We then introduce an HI approach that fosters self-reflection on values by connecting value classification and estimation. We conclude by describing several applications that are being developed based on value inference, ranging from support for deliberative policy-making to behavior change for diabetes patients.
Paper presented: Enrico Liscio et al. "Value Inference in Sociotechnical Systems", AAMAS '23 (Blue Sky Ideas Track), London, United Kingdom, IFAAMAS, 1774-1780.
When: 6 Nov 2023, 15:00 CET
Speaker: Marc Serramia Amoros, City University of London, UK
Title: Smart Participatory Budgets: Our Experience with Decidim Barcelona
Abstract: Each year, governments and organisations hold thousands of participatory budget processes, but despite their extensive adoption they still present many challenges. These include low participation numbers, uneven representation, lack of debate, or suboptimal budget allocations. Importantly, while there has been extensive research in AI targeting these problems, there is still a gap between theory and practice. We approached the participatory budget organiser of our city, Decidim Barcelona, to help bridge this gap. Hence, we met with them to understand their problematics and propose AI solutions for them. We focused on improving the quality of debates and of the budget allocation solution. First, we designed an argumentation-based debate platform that encouraged participation. Second, we detected the budget allocation process used by Decidim was suboptimal in terms of citizen satisfaction and proposed an improvement. Our experience with Decidim while positive was only a small step towards implementing AI research in practice, currently we are aiming at bigger challenges. We are studying new AI approaches to tackle low participation numbers and uneven representation with the aim of both considering the voice of those who did not participate, and encourage governments to rely more on participatory processes.
Papers presented:
Marc Serramia et al. "Citizen Support Aggregation Methods for Participatory Platforms", Artificial Intelligence Research and Development. IOS Press, 2019. 9-18.
Marc Serramia et al. "Optimising Participatory Budget Allocation: The Decidim Use Case", Artificial Intelligence Research and Development. IOS Press, 2019. 193-202.
When: 29 June 2023, 14:00 CEST
Speaker: Serena Villata, CNRS, France
Title: Mining Natural Language Arguments from Text: Achieved Results and Open Challenges