Solar activity, space weather and human health: can ChatGPT assist systematic literature reviews?

Research linking solar phenomena to health conditions has been a subject of interest in the scientific community for nearly a century. This systematic literature review seeks to deliver a novel, transparent, and methodical examination of the literature on the subject, highlighting potential research gaps, particularly concerning the employed data, data sources, prospective novel research avenues, and the current status of the field. This systematic literature review employs ChatGPT for data extraction from the gathered papers and offers guidelines and a potential framework for integrating such models into systematic literature reviews. The findings indicate that prospective research directions should focus on social media data, particularly regarding health and human-related information, as solar phenomena data is readily available through various online services. The majority of studies focus on geomagnetic activity or geomagnetic storms and cardiovascular conditions, while a minority address psychological states. ChatGPT supplied valuable information that was verified using four anchor hallucination questions and additional assessments through data analysis. The significant reduction in the time required to extract data from research paper databases is the primary advantage of employing large language models, enabling researchers to analyze a larger quantify of papers, provided the model’s prompt is carefully created with anchor hallucination questions and thorough validation by the researchers.