Serendipitous Language Learning in Mixed Reality

Mixed Reality promises a new way to learn in the world: blending holograms with our surroundings to create contextually rich experiences that find their place in our daily routine. Existing situated learning platforms limit the learner and implicitly enforce the designer's intent by hard-coding content and predetermining what elements in the environment are actionable. In this paper, we define the framework of Serendipitous Learning in Mixed Reality as situated, incidental learning that occurs naturally in the user's environment and stems from the learner's curiosity. This framework is explored within the context of second language acquisition. We present WordSense, a Mixed Reality platform that recognizes objects in the learner's vicinity and embeds holographic content that identifies the corresponding word, provides sentence and definition cues for practice, and displays dynamic audiovisual content that shows example usage. By employing markerless tracking and dynamically linked content, WordSense enables serendipitous language learning in the wild.