Abstract

Fosca Giannotti
Titolo: Human-centered Artificial Intelligence.

Abstract: The future of AI lies in enabling people to collaborate with machines to solve complex problems. Like any efficient collaboration, this requires good communication, trust, clarity, and understanding. Explaining to humans how AI reasons is only a part of the problem: we also must be able to design AI systems that understand and collaborate with humans. Hybrid decision-making systems aim at leveraging the strengths of both human and machine agents to overcome the limitations that arise when either agent operates in isolation.

The lecture will highlight the steps needed for promoting human-AI collaboration and seamless interaction maintaining the human responsibility of choice through a progressive disclosure to prevent cognitive overload. Three distinct paradigms, characterized by a different degree of human agency and machine autonomy will be discussed: i) human oversight, with a human expert monitoring AI prediction augmented by explanation; ii) Learning to defer, in which the machine learning model is given the possibility to abstain from making a prediction when it receives an instance where the risk of making a misprediction is too high; iii) collaborative and interactive learning, in which human and AI engage in communication to integrate their distinct knowledge and facilitate the human ability to make informed decisions.

Dirk Hovy
Title: Understanding society through text – computational linguistics

Abstract: Text is an important component in today’s information-rich society, and provides signal for a variety of social studies. Participants will learn about text representations and how to use and interpret modern embedding models effectively. The course will also cover text classification techniques and clustering, which are critical for understanding complex patterns in language data. We will finally cover geocoding, which is a method for incorporating geographical context into text-based analysis. As running example, we will retrace the main components of the paper “Capturing Regional Variation with Distributed Place Representations and Geographic Retrofitting”, both in theory and practice (via Jupyter notebooks).

This workshop provides a comprehensive practical guide for social scientists interested in using Python to dissect and comprehend text data.

David Lazer
Title: Meaningful computational social science

Abstract: This presentation will discuss the trajectory of the field, with a particular focus on the fusion of computational and social science methods. The scientific opportunity of computational social science is to understand/predict/describe individual and societal outcomes, incorporating context, content, and connectivity as interacting factors. However, the core social science concerns of developing meaningful measures and evaluating generalizability of findings are often especially problematic in CSS research. The objective of this lecture will be to connect the dots between method and insight for the field.

Filippo Menczer
Title:Computational social science methods to study online virality and its manipulation

Abstract: As social media have become major channels for the diffusion of news and information, it is critical to understand how the complex interplay between cognitive, social, and algorithmic biases triggered by our reliance on online social networks makes us vulnerable to manipulation and disinformation. We focus on two key factors that contribute to online virality: the structure of the social network and the engagement mechanisms that manage our limited attention. This lecture overviews computational social science efforts (spanning network analytics, modeling, and machine learning) to study the viral spread of misinformation and to develop tools for countering the online manipulation of opinions.

Alexandra Olteanu
Title: Fairness, Accountability, Transparency, and Ethics

Abstract: The Responsible AI field—often also referred to as FATE (Fairness, Accountability, Transparency, and Ethics)—has seen a rapid growth over the last decade, primarily driven by growing concerns about how computational systems can exacerbate, replicate, or give rise to harms.

In this lecture, we will overview foundational concepts, problem formulations, and existing tensions, as well as discuss how the field has evolved over time and why. We will review Responsible AI work through the lens of data, measurement, values, human-AI interactions, and people. In doing so, we will also draw connections to early work in the computational social science and social computing communities.

Dino Pedreschi
Title: Social Artificial Intelligence: Challenges of the Human-AI Ecosystem

Abstract: The rise of large-scale socio-technical systems in which humans interact with AI systems (including assistants and recommenders) multiplies the opportunity for the emergence of collective phenomena and tipping points, with unexpected, possibly unintended, consequences. This is apparent even in such simple everyday applications as navigation systems where suggestions may create chaos if too many drivers are directed on the same route. Similarly personalised recommendations on social media may amplify polarisation, filter bubbles, and radicalisation. On the other hand, we may learn how to foster “wisdom of crowds” and collective action effects to face social and environmental challenges. In order to understand the impact of AI on socio-technical systems and design next-generation AIs that team with humans to help overcome societal problems rather than exacerbate them, we propose to build the foundations of Social AI at the intersection of Complex Systems, Network Science and AI, and discuss the open challenges of human-AI ecosystems such as social media, car navigation systems, and generative AI platforms, outlining possible research avenues.

Alessandro Vespignani
Title: Multiscale Computational Models in Infectious Disease Epidemiology for Policy Decision-Making

Abstract: The field of epidemic modeling has undergone significant advancements, yet it faces ongoing challenges like data quality, accessibility, and the complexities of integrating diverse data sources. These problems not only underscore the dynamic nature of the field but also position it as a rapidly evolving and exciting frontier in scientific research. This course offers a comprehensive exploration of the latest developments in infectious disease modeling, focusing on integrating contact patterns and socio-behavioral data. Participants will learn how these models can provide timely assessments of intervention strategies and essential situational awareness, particularly in scenarios with limited data. Drawing from lessons learned during recent pandemics, the course will delve into the evolving role of data-driven policymaking in infectious disease management. This will include discussions of the most recent developments in  multi-model ensembles and “hub” approaches for forecasting, scenario modeling, and effective public health communication.