On September 25th, 2023 at 1.30 pm Alessandro Bozzon, Delft University of Technology, will hold a seminar on "From Annotation to Contestation: Crowd Computing for AI-powered Systems" in DEIB Alpha Room.

Crowd Computing is a computational paradigm for human-centered intelligent systems that advocates the adoption of human intelligence at scale to improve artificial intelligence systems' performance, especially concerning robustness, interpretability, usability, and trustworthiness. While historically linked to large-scale data generation, annotation, and labeling tasks — most famously with the ImageNet dataset — its application has rapidly expanded to encompass real-time human-in-the-loop applications, such as content moderation and continuous performance evaluation. In this talk, I will present the results of our research focused on opening the black box of modern machine-learning systems with Crowd Computing. I will present recent work on crowd-enabled 1) interpretation and 2) debugging of computer vision systems. Finally, I will introduce the topic of Contestable AI and discuss how crowd-powered approaches could increase AI systems' transparency by making them open and responsive to human intervention throughout their lifecycle.

Alessandro Bozzon is Professor of Human-Centered Artificial Intelligence with the Knowledge and Intelligence Design (KInD) section, Department of Sustainable Design Engineering (SDE) of the Faculty of Industrial Design Engineering (IDE); and a part-time professor with Web Information Systems section, Department of Software Technology of the Faculty of Electrical Engineering, Mathematics, and Computer Science (EEMCS) of Delft University of Technology. As of November 2020, he serves as Head of the Department of Sustainable Design Engineering. His research lies at the intersection of human-computer interaction, human computation, user modelling, and machine learning. He is interested in developing methods and tools that support the design, development, control, and operation of AI-enabled systems that are well-situated around actual human characteristics, values, intentions, and behaviors. By investigating the relationship between the science and practice of design, and the digital technology that fuels intelligent products, services, and systems, he studies and builds novel Human-Centered Artificial Intelligence methods and tools that combine the cognitive and reasoning abilities of (groups of) individuals, with the computational powers of machines, and insights from large amount of heterogeneous data. Alessandro has published more than 150 papers in peer-reviewed international journals (e.g. VLDBJ, ACM TWEB, IEEE IC, IEEE Access) and conferences (e.g. CHI, CSCW, HCOMP, WWW/TheWebConference, AAAI, IJCAI, RecSys, ESWC, ISWC)). In 2017 he received the IBM Faculty Award for his work on Enterprise Crowd Computing. He currently serves as a member of the steering committee of the International Conference of Web Engineering (ICWE) conference series, and as Associate Editor of the International Journal of Web Engineering (JWE). He regularly serves as Program Chair or Area Chair for major international conferences.