An AI academic advising digital human model persona framework

Since the emergence of generative AI, experimentation with AI tools for academic advising has increased. Social science fields are called upon to inform more human values-sensitive design approaches. A digital human model for academic advising provides a creative methodological space for mapping the rich human dynamics of academic advising with technology practices and could be used in the design of AI advising systems. This paper proposes, as the first steps towards a digital human model, a Dialogic-Relational Translation device and a related Persona framework, building on literature and the academic advising practices of 19 academic advisors at a Flemish university. The research combines a relational view of technology with dialogic educational theory to create a Dialogic-Relational translation device through thematic analysis of semi-structured interview transcripts and an interative process of zooming and refocusing. The translation device is then applied to the academic advising context to develop a Dialogic-Relational AI-System Persona framework with four dynamic academic advising positionalities — Consultant, Coach, Counsellor, and Collaborator. These instruments are proposed as groundwork for a conceptual and empirical approach to analysing technology-related advising practices. The study demonstrates that although there are no one-dimensional solutions to designing models for human-machine (advisor-AI) entangled practices, the Dialogic-Relational Translation device and Persona framework can support a more nuanced institutional understanding and inform the development and analysis of AI academic advising systems.