1 Humboldt University Berlin (GERMANY)
2 University of Applied Sciences (GERMANY)
3 St. Gallen University of Teacher Education (SWITZERLAND)
About this paper:
Appears in: INTED2020 Proceedings
Publication year: 2020
Pages: 9313-9320
ISBN: 978-84-09-17939-8
ISSN: 2340-1079
doi: 10.21125/inted.2020.2576
Conference name: 14th International Technology, Education and Development Conference
Dates: 2-4 March, 2020
Location: Valencia, Spain
Recent advances in sensor technology, especially wearable sensors, allow researchers in various field to investigate the physiological, environmental and psychological states of people. The investigated inference is, then, used to provide an individual with an awareness of one’s state in visual, auditory and haptic form. For example, wearable watch which mainly includes PPG (Photoplethysmogram), EDA (Electrodermal Activity) and Accelerometer detects one’s physiological and psychological states. Based on the detected state, users are alerted with a chime or beep along with an abrupt vibration, followed by a visual representation of their physiological and emotional states.

In an adaptive personalized learning context, a wearable sensor can serve as a mean to study learners’ biological and environmental signals to infer their cognitive and affective states. However, using a commercially available wearable sensor for this purpose is not a solution even though it presents a convenient and attractive means for investigation. Selecting sensors that detect cognitive and affective state of learners should stem from understanding of learners and a learning context and data privacy and ethics that are commonly overlooked in a commercially available solution should be adhered.

The obtained individualized data can then be used to derive learners’ state and environment. Once learners’ cognitive and affective states are detected, learning contents (e.g., difficulty levels or topics) can be adapted accordingly and learners can be provided with personalized feedback and messages.

In this paper, we present two cases of utilizing a wearable sensor as a mean to support learners’ emotion. In the first case study, we:
1) describe our motivation and approaches on selecting and using sensors to detect behavioural, cognitive, and, affective states of learners
2) discuss our design approaches focusing on HCI (Human Computer Interaction) to provide learners with external regulation of their emotion.

In the second case study, we present a use-case of a wearable sensor in a Kindergarten to detect digital anxiety and support them in emotionally difficult situations.

This paper presents approaches, methods and first results for emotion detection from physiological sensor data obtained in the studies. Then we show how sensor data analysis can support awareness and self-regulation of a learner and we present first results on emotional support for Kindergarten children using sensor data.

Based on our experience in utilizing a wearable sensor in a learning context, we share our future study plan in a specific learning and teaching scenario (e.g. sports study, social robot) and in a large-scale classroom.
sensor based learning, wearable, self-regulated learning, external regulation, learning companion.