Scholarlink

<p>The rapid advancement of generative artificial intelligence (GenAI), particularly large language models (LLMs), has introduced transformative possibilities for learning analytics (LA) in higher education. Despite growing empirical work, no prior review appears to provide an integrated synthesis of GenAI studies in higher education learning analytics through the combined lens of functional roles and decision support applications. Existing reviews address broader AI-in-education topics or isolated learning analytics functions rather than providing an integrated synthesis of these dimensions. This paper presents a systematic review of 38 studies, following the PRISMA 2020 guidelines, to address four research questions: (RQ1) what GenAI techniques and models are used in higher education LA research; (RQ2) what functional roles GenAI performs across LA processes; (RQ3) what decision support applications are enabled; and (RQ4) what research gaps and future directions exist. Drawing on studies retrieved from Scopus, Web of Science, and IEEE databases, supplemented by citation snowballing, we propose a taxonomy of six functional roles and identify nine categories of decision support applications. GPT-4 and related models emerged as the dominant technologies, while multimodal LLMs, retrieval-augmented generation, and generative adversarial networks represent emerging AI directions. The review reveals critical gaps in scalability, equity, interpretability, longitudinal evaluation, and ethical governance. These findings provide a foundation for researchers, educators, and institutional decision-makers seeking to leverage GenAI in educational practice.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A systematic review of generative AI in higher education learning analytics and decision support

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月2日

查看全文
  • Husna Hafiza Razami,
  • Abtar Darshan Singh,
  • Gabriel Avelino R. Sampedro,
  • Jonathan Jeevan Strivinas Kovilpillai,
  • Fahd Ali Raza,
  • Rabia Anwar,
  • Analisa Hamdan,
  • Fumiko Konno

摘要

The rapid advancement of generative artificial intelligence (GenAI), particularly large language models (LLMs), has introduced transformative possibilities for learning analytics (LA) in higher education. Despite growing empirical work, no prior review appears to provide an integrated synthesis of GenAI studies in higher education learning analytics through the combined lens of functional roles and decision support applications. Existing reviews address broader AI-in-education topics or isolated learning analytics functions rather than providing an integrated synthesis of these dimensions. This paper presents a systematic review of 38 studies, following the PRISMA 2020 guidelines, to address four research questions: (RQ1) what GenAI techniques and models are used in higher education LA research; (RQ2) what functional roles GenAI performs across LA processes; (RQ3) what decision support applications are enabled; and (RQ4) what research gaps and future directions exist. Drawing on studies retrieved from Scopus, Web of Science, and IEEE databases, supplemented by citation snowballing, we propose a taxonomy of six functional roles and identify nine categories of decision support applications. GPT-4 and related models emerged as the dominant technologies, while multimodal LLMs, retrieval-augmented generation, and generative adversarial networks represent emerging AI directions. The review reveals critical gaps in scalability, equity, interpretability, longitudinal evaluation, and ethical governance. These findings provide a foundation for researchers, educators, and institutional decision-makers seeking to leverage GenAI in educational practice.