A Sentiment Analysis of Data on Outings from Mothers with Infants and Toddlers during the COVID-19 Pandemic

Global travel demand has been severely suppressed worldwide since the outbreak of the COVID-19 pandemic. In this study, we conducted a sentiment analysis of data on outings provided by mothers with infants and toddlers before and after the COVID-19 pandemic. These data were expressed in Japanese and consisted of log data from a smartphone application called mamari in which users communicate to solve and share questions and concerns about pregnancy, childbirth, and childcare. A sentiment classifier was trained using an annotated list of positive and negative sentiment words and a pretrained word embedding. Then, we calculated the mean sentiment analysis scores in the question sentences and examined the relationship between the sentiment scores of the questions and the daily confirmed cases of COVID-19 in Japan. The results show that the sentiment analysis values for the question data decreased with increasing numbers of confirmed cases after the COVID-19 pandemic began.