Ethical Considerations for Artificial Intelligence Tools in Academic Research and Manuscript Preparation: A Web Content Analysis

Abstract

Abstract Our research investigated the potential effects of incorporating artificial intelligence (AI) techniques into scholarly publications, specifically big language models. The study employs a qualitative methodology and web content analysis to understand various publishers' guidelines. It examines the potential applications and outcomes of AI in publishers. The analysis has revealed insightful findings regarding the use and implications of AI tools in academic publishing. Agglomeration analysis has uncovered distinct clusters of terms, indicating semantic relationships and thematic cohesion within the dataset. Notably, ‘Large’ and ‘Models’ have formed a coherent cluster, highlighting the significance of large-scale language models in scholarly discourse. Similarly, factor analysis has identified thematic clusters related to AI usage, emphasising aspects such as accuracy, responsibility and the role of authors in AI-assisted work. Semantic mapping has further elucidated thematic dimensions, highlighting linguistic frameworks, work-related constructs, methodological frameworks, AI technologies and publication dynamics. Evaluation metrics have consistently demonstrated cohesion, coherence and lexical diversity across varying numbers of topics, underscoring the robustness of the semantic mapping approach. Additionally, the Silhouette coefficient has provided insights into cluster quality, indicating strong cohesion within specific thematic clusters while hinting at potential overlaps in others. Co-occurrence matrix and cross-tabulation analysis have revealed association and frequency distribution patterns among terms, shedding light on prevalent themes and topics within the dataset. Finally, the proximity plot has illustrated the strength of associations between keywords and accuracy, emphasising central themes and moderate thematic relevance.

Ähnliche Arbeiten