An Educational Conversational Agent for GDPR

Abstract

Large-scale learning scenarios as well as the ongoing pandemic situation underline the importance of educational technology in order to support scalability and spatial as well as temporal flexibility in all kinds of learning and teaching settings. Educational conversational agents build on a long research tradition in intelligent tutoring systems and other adaptive learning technologies but build for interaction on the more recent interaction paradigm of conversational interaction. In this paper, we describe a tutorial conversational agent, called GDPRAgent, which teaches a lesson on the European General Data Protection Regulation (GDPR). This regulation governs how personal data must be treated in Europe. Instructionally, the agent’s dialogue structure follows a basic GDPR curriculum and uses Bloom’s revised taxonomy of learning objectives in order to teach GDPR topics. This overall design of the dialogue structure allows inserting more specific adaptive tutorial strategies. From a learner perspective, the learners experience a completely one-on-one tutorial session in which they receive relevant content (is “being taught”) as well as experiences active learning parts such as doing quizzes or summarising content. Our prototype, therefore, illustrates a move away from the dichotomy between content and the activity of teaching/learning in educational technology.