Quality Evaluation of Online Mental Health Education Based on Reinforcement Learning in the Pandemic

Abstract

The COVID-19 pandemic has become one of the biggest major health crises reported due its massive impact on many countries. From mental health experts, we know that we cannot lose sight of an equally alarming issue which is the long-term mental health impact the pandemic is going to leave on the society. The rapid spread of the pandemic gives little chance to prepare for or even process all that has happened in terms of job losses and the complete uprooting of everyday life and relationships. It is understandable that students may feel irritable, frustrated, or sad sometimes. Loneliness, confusion, and anxiety are also common, but the issue is how we can know if students’ emotions are a normal reaction to an abnormal situation. Therefore, online mental health education has become pretty important for students during the pandemic. Furthermore, it is important to evaluate the quality of online mental health education through microlessons. In this paper, based on Q-learning algorithm, the real-time adaptive bitrate (ABR) configuration parameters mechanism is proposed to detect the changes of network state constantly and select the optimal precalculated configuration according to the current network state. The simulation results show that the proposed algorithm based on Q-learning outperforms other baselines in average latency, average bitrate, and Mean Opinion Score (MOS) on Chrome DevTools and Clumsy. Meanwhile, the experimental results also reveal that the average number of identified mental health problems of the proposed mechanism has always been the best with the bandwidth from 10 Mbit/s to 500 Mbit/s.