Applications of artificial intelligence in developing systematic literature reviews: A systematic literature review using semantic analysis

Abstract

This systematic literature review investigates how artificial intelligence techniques are used to support systematic literature reviews in computer science and artificial intelligence research, with emphasis on semantic analysis, automated screening, topic discovery, knowledge extraction, and conceptual synthesis.

The manuscript follows the PRISMA 2020 logic and proposes a transparent search strategy across Scopus, Web of Science, IEEE Xplore, ACM Digital Library, ScienceDirect, SpringerLink, and ERIC where educational computing is relevant. A semantic-analysis protocol was designed to extract recurring concepts, keywords, relationships, and thematic patterns from the selected studies.

The synthesis organizes the literature into six themes: AI-assisted search and retrieval, semantic screening and classification, topic modelling and clustering, knowledge extraction and ontology construction, quality appraisal and bias detection, and human-AI collaboration in evidence synthesis.

AI can enhance speed, breadth, and analytical depth in SLRs, yet publication-ready reviews still require human judgment, transparent reporting, validation of AI outputs, and ethical documentation of tool use.