Work-life balance is a critical aspect of well-being, and its disruption can lead to various negative outcomes, including depression. It has been estimated that many people extend work activities beyond their stipulated work hours in Nigeria, also bearing the impact of the steady economic challenges, its effect on the living conditions of an average citizen, these conditions alone may lead to several forms of depression and work-life balanced induced depression is an example. Empathy, both from humans and machines, plays significant roles in supporting individuals experiencing work-life balance-induced depression. This research develops a large language model that compares human and machine empathy in addressing work-life balance-induced depression. The study employs a mixed-methods approach, incorporating both quantitative and qualitative data. A comprehensive review provides a foundation for understanding the underlying factors and potential interventions related to work-life balance-induced depression. Then, a large language model is developed from fine-tuning existing GPT-3 davinci model, leveraging advanced natural language processing techniques. The model is trained on a dataset comprising real-life scenarios related to work-life balance-induced depression, personal experiences, challenges, and coping mechanisms. Both human participants and the large language model are presented with these scenarios and asked to provide empathetic responses. Evaluation of the model’s performance on the generated empathetic statements involved the use of perplexity, accuracy and F1-score as metrics. The findings of this research contribute to a better understanding of the potential role of machine empathy in addressing work-life balance-induced depression.

Keywords: work-life balance, empathy, depression, large language model, predictive model