Leveraging Artificial Intelligence for Systematic Reviews: The FRAISR Reporting Framework and guidance for researchers

Abstract

Systematic Reviews (SR) are crucial in synthesizing research findings, yet they are resource-intensive and prone to errors in critical steps like screening and data extraction. Artificial Intelligence (AI) offers potential in conducting SRs by decreasing time needed, reducing errors, and streamlining processes. This paper proposes a novel AI Guidance and Reporting Framework for Systematic Reviews, addressing the challenges inherent in traditional SR methodologies and the opaque nature of AI models. Existing frameworks for SRs are discussed, and the proposed AI framework, and its usage in a machine-readable format presented. Key challenges and attempts to use AI in each stage of a SR are highlighted, including known limitations to guide future research. By proposing this framework, it is intended to standardize AI application in SRs, fostering ethical and efficient research practices and advancing the field towards more robust and transparent methodologies.

Ähnliche Arbeiten