Building a Psychological Ground Truth Dataset with Empathy and Theory-of-Mind During the COVID-19 Pandemic

This paper is a summary of the first research project that I was involved during my Master’s. We tried to understand how Koreans are going through COVID-19, and construct a ground-truth dataset of social cognition and emotion that constitues of both journal entries and questionnaire responses. In this paper, we analyzed 19,025 journal enteries of daily emotional experiences of 3,805 Korean residents, along with the psychological ground-truth values of the journal writers, collected from October to December 2020. Participants varied in their ages from the early 20s to late 80s and had various social and economic statuses. We used topic modeling and K-means clustering on the Word2Vec vectors trained on the journal entries, and found that pandemic impacted the majority of daily lives, and participants often reported negative emotional experiences. The most frequent topics included: responses to confirmed cases, health concerns of family members, anger towards people without masks, stress-relief strategies, change of the lifestyle, and preventive practices. This dataset will help portray the psychological characteristics of Koreans during COVID-19, and serve as benchmark data for large-scale and computational methods for identifying mental health levels based on text.

The full dataset, completed after this paper, contains approximately 7,000 responses on psychological variables and 30,000 diary entries. For the full access to the dataset, visit here.