Guest Talk: "Theory-driven dialogue design for reflective conversational agents" by Assoc.-Prof. Dr. Viktoria Pammer-Schindler
News from Jan 28, 2021
LANGUAGE: English PRESENTATION SLIDES: Slideshare
Imagine an intelligent entity that helps YOU to reflect - on yourself and your goals, on your experiences, on how to connect newly received knowledge to your existing ways of thinking. This is of course what good human teachers, coaches, and mentors can do. Can computers do that, too?
Research is ongoing on various issues of how to enable conversational agents - computer software endowed with some semblance of intelligence - to facilitate just such open-ended, reflective dialogues. This capability of conversational agents to guide through reflection by leading a reflective conversation is a different capability than seeming human, being able to induce human users to feel affect, or teaching factual knowledge (which is important, too!).
In this talk, a blueprint dialogue structure for reflecting on a single experience with the goal to learn for the future, will be described. This dialogue structure has been developed based on reflection theory and two use cases in work-related learning.
Further, the concept of reflection scripts will be introduced: This is a framework for designers of reflective technology to think in a structured manner about what type of reflection is the goal of design and how to structure the connection between reflection and other human activities (macro scripts), and how to structure guidance through such reflection (micro scripts).
Viktoria Pammer-Schindler is an associate professor at Graz University of Technology and area head at the non-for-profit research organisation Know-Center. Viktoria researches working and learning in the ongoing digital transformation; and her research sits at the intersection of technology-enhanced learning and learning sciences, human-computer interaction, and information systems research.
Viktoria's approach to research is to go the full cycle, from understanding working and learning based on empirical work, to designing socio-technical interventions, to evaluating these in the wild. I use methods from user-centered and contextual design; and carry out experimental studies, design studies and field studies. Analysis methods are qualitative, quantitative, and data analytics based when log data can be used; and I also study participatory design and similar methods as methods in innovation. Technologies are interactive, adaptive, and use data analytics and artificial intelligence.
Viktoria researches reflective technology and reflection analytics; designing for contextualized learning about AI, designing for knowledge-intensive activities in particular where they relate to knowledge construction and organisational learning; designing for knowledge construction, reflection and decision-making in innovation activities; and overall the inter-relationships between working, learning and technology in the ongoing digital transformation.