Large language models (LLMs), such as OpenAI's Generative Pre-trained Transformer (GPT), are a class of language models that have demonstrated outstanding performance across a range of natural language processing (NLP) tasks. LLMs have become a highly sought-after research area because of their ability to generate human-like language and their potential to revolutionize science and technology. In this study, we conduct bibliometric and discourse analyses of scholarly literature on LLMs. Synthesizing over 5,000 publications, this article serves as a roadmap for researchers, practitioners, and policymakers to navigate the current landscape of LLMs research. We present the research trends from 2017 to early 2023, identifying patterns in research paradigms and collaborations. We start with analyzing the core algorithm developments and NLP tasks that are fundamental in LLMs research. We then investigate the applications of LLMs in various fields and domains, including medicine, engineering, social science, and humanities. Our review also reveals the dynamic, fast-paced evolution of LLMs research. Overall, this article offers valuable insights into the current state, impact, and potential of LLMs research and its applications.

Background: Prior studies indicate that older members of LGBTQ+ communities have specific health provision and health information needs related to coping with COVID-19, its long-term effects, and the social and economic impact of the pandemic. This study addresses the issue of a lack of timely, complete, and high-quality data about this population’s healthcare and healthcare information needs and behaviors. Recognizing also that this is a diverse population made up of multiple communities and identities with different concerns and experiences, this research seeks to develop and refine a method that can provide additional nuanced data and insights that can support improved and more closely targeted health interventions and online information provision.Methods: We use computational discourse analysis, which is based on NLP algorithms, to build and analyze a digital corpus of online search results containing rich, wide-ranging content such as quotes and anecdotes from older members of LGBTQ+ communities as well as practitioners, advice, and recommendations from policymakers and healthcare experts, and research outcomes. In our analysis, we develop and apply an innovative disparities and resilience (D&R) framework to identify external and internal perspectives and understand better disparities and resilience as they pertain to this population.Results: Results of this initial study support previous research that LGBTQ+ elders experience aggravated health and related social-economic disparities in comparison to the general population of older people. We also find that LGBTQ+ elders leverage individual toughness and community closeness, and quickly adapt mentally and technologically, despite inadequate social infrastructure for sharing health information and elders’ often low social economic status. The methods used therefore are able to surface distinctive resilience in the face of distinctive disparities.Conclusions: Our study provides evidence that methodological innovation in gathering and analyzing digital data relating to overlooked, disparately affected, and socially and economically marginalized intersectional communities such as LGBTQ+ elders can result in increased external and self-knowledge of these populations. Specifically, it demonstrates the potential of computational discourse analysis to surface hidden and emerging issues and trends relating to a multi-faceted population that has important concerns about public exposure in highly timely and automated ways. It also points to the potential benefits of triangulating data gathered through this approach with data gathered through more traditional mechanisms such as surveys and interviews.Trial registration: Not Applicable.

This article reports on a study using machine learning to identify incidences and shifting dynamics of hate speech in social media archives. To better cope with the archival processing need for such large-scale and fast evolving archives, we propose the Data-driven and Circulating Archival Processing (DCAP) method. As a proof-of-concept, our study focuses on an English language Twitter archive relating to COVID-19: Tweets were repeatedly scraped between February and June 2020, ingested and aggregated within the COVID-19 Hate Speech Twitter Archive (CHSTA), and analyzed for hate speech using the Generative Adversarial Network–inspired DCAP method. Outcomes suggest that it is possible to use machine learning and data analytics to surface and substantiate trends from CHSTA and similar social media archives that could provide immediately useful knowledge for crisis response, in controversial situations, or for public policy development, as well as for subsequent historical analysis. The approach shows potential for integrating multiple aspects of the archival workflow and supporting automatic iterative redescription and reappraisal activities in ways that make them more accountable and more rapidly responsive to changing societal interests and unfolding developments.