Computer and Information Technology Seminar

Helping Older Users with their Smartphone: The Role of Social Support

Older adults often experience a range of challenges when using smartphones, adding to a growing digital divide between generations and limiting access to a wide range of socio-economic interactions (e.g., healthcare, socializing). One major way to overcome these challenges is social support, that is, assistance from family, friends, and other social peers through interpersonal communication. The goal of the present study is to provide insights into how social support can help older users in overcoming their challenges with smartphone usage. To this end, we conducted an online survey with social peers who have helped older users in their network with their smartphone usage. Based on cluster analysis, we identify three main helper communication types, each of which perceive and approach their interactions with older smartphone users in a different way. Our findings provide important insights into how social peers can help older users with developing understanding and building independence.

BadFU: Backdoor Federated Learning through Adversarial Machine Unlearning

Federated learning (FL) has been widely adopted as a decentralized training paradigm that enables multiple clients to collaboratively learn a shared model without exposing their local data. As concerns over data privacy and regulatory compliance grow, machine unlearning, which aims to remove the influence of specific data from trained models, has become increasingly important in the federated setting to meet legal, ethical, or user-driven demands. However, integrating unlearning into FL introduces new challenges and raises largely unexplored security risks. In particular, adversaries may exploit the unlearning process to compromise the integrity of the global model. In this paper, we present the first backdoor attack in the context of federated unlearning, demonstrating that an adversary can inject backdoors into the global model through seemingly legitimate unlearning requests. Specifically, we propose BadFU, an attack strategy where a malicious client uses both backdoor and camouflage samples to train the global model normally during the federated training process. Once the client requests unlearning of the camouflage samples, the global model transitions into a backdoored state. Extensive experiments under various FL frameworks and unlearning strategies validate the effectiveness of BadFU, revealing a critical vulnerability in current federated unlearning practices and underscoring the urgent need for more secure and robust federated unlearning mechanisms.

From Principles to Practice: A Deep Dive into AI Ethics and Regulations

This talk thoroughly analyzes the ground-breaking AI regulatory framework proposed by the European Union. It delves into the fundamental ethical principles of safety, transparency, non-discrimination, traceability, and environmental sustainability for AI developments and deployments. Considering the technical efforts and strategies undertaken by academics and industry to uphold these principles, we explore the synergies and conflicts among the five ethical principles. Through this lens, work presents a forward-looking perspective on the future of AI regulations, advocating for a harmonized approach that safeguards societal values while encouraging technological advancement.

AI psychosis: Where do the responsibilities lie?

Media coverage of select cases of AI psychosis has raised questions about AI regulation, technology design, legal accountability, and support for those who are at-risk of developing psychosis (or triggering an acute episode) when interacting with chatbots or other AI platforms. This presentation will examine case studies to demonstrate that, in the next digital revolution, questions of ethics, psychology, legal frameworks, and cross-disciplinary collaborations will become more important than ever.

Applied Computing Research in Education: A review of two case studies

In this talk, I will speak on my experience of bringing applied computing research in education to develop research projects in educational technology and technology-enhanced learning fields. Leveraging expertise in Computing and IT while also being an academic involved in teaching (i.e. being an end-user and stakeholder), I had the opportunity to develop some innovative research projects integrating these domains. These projects led to PhD projects/supervisions, publications and sometimes impact beyond publications. In this talk, I will present two case studies of projects in Applied Computing in Education along with their outcomes.

In the first case study, an innovative lab and related artefacts were developed. Also, lab activities were re-designed based on educational theories. Next, real-world evaluations of the labs were undertaken to evaluate the impact of these innovations.  In the second case study, an artefact was created using both educational theories and machine learning for Learning Analytics Interventions (LAIs). Evaluation included deploying these artefacts in multiple real-world courses and also multiple offerings of the same course. I will delve deeper into ethics and quasi-experimental methods used. Outcomes included publications, PhD completions, and lab activities that are still in use today.

The emergence of Artificial Intelligence (AI) brings about many opportunities to conduct research in these fields. Finally, I will outline a set of basic steps that I use in developing a research project (i.e. a potential PhD student project) in applied computing in education.

TBD

TBD

TBD

TBD