Опубликована: Авг. 28, 2023

To make social robots effective in education, they need to be autonomous both terms of assessing the student's engagement state as well intervening effectively soft real-time when necessary. Hidden Markov Model (HMM) is an interpretable machine learning technique for modeling temporal data that commonly used post-hoc analyse latent processes. In this paper, we contribute by proposing HMM-based intervention methodology and classifying student either productive or unproductive real-time. The system identifies tracks states patterns not conducive learning, a robot triggered whenever too-high non-productive detected. pilot study with 22 children, evaluate 1) effectiveness interventions on students' gains behaviors found 2) perception robotic interventions. Results suggest have positive effect post-test scores relative baseline robot, although there isn't significant difference gains. Moreover, try induce reflective are most inducing required behavior, followed communication-inducing Lastly, usefulness does reflect their actual effectiveness.

Язык: Английский