PhD Dissertation Proposal by ZHANG Tianyi | Designing Context-Aware AI Intervention for Interpersonal Communication

ABOUT THE TALK

Conversational agents are increasingly evolving from tools that assist individual users into systems that actively participate in communication between people. Unlike traditional human–AI interaction, interpersonal communication requires AI systems to reason not only about individual users but also about the social, relational, and epistemic contexts in which communication takes place. However, existing research provides limited understanding of how AI interventions should be designed and adapted across different communication contexts.

This dissertation investigates how context-aware AI systems can effectively intervene in and support interpersonal communication. It addresses two complementary research problems: (1) establishing a systematic design space for AI interventions and (2) understanding how intervention strategies should adapt to different communication contexts. To address the first problem, the dissertation begins with a systematic literature review of group conversational agents, synthesizing prior work into a conceptual design space of interaction challenges, agent roles, and intervention strategies. Building on this foundation, it examines AI intervention in two contrasting contexts: socially intimate communication, through a scenario-based design study of intergenerational communication, and professional communication, through an ongoing Wizard-of-Oz prototype study of interdisciplinary collaboration.

Collectively, these studies establish a conceptual design space for AI interventions, provide empirical insights into how expectations of AI intervention vary across communication contexts, and generate design implications for developing context-aware conversational agents that effectively support interpersonal communication.

ABOUT THE SPEAKER

ZHANG Tianyi is a PhD candidate in Computer Science at the School of Computing and Information Systems, Singapore Management University, under the supervision of Associate Professor Tony TANG and Assistant Professor Li Jiannan. Her research primarily focuses on Human–AI Interaction, particularly the design of AI systems that support interpersonal and group communication.