Lifelong learning research in WOS and CNKI: A comparative topic modeling analysis

Abstract Understanding and practicing lifelong learning is the focal point of community education and continuing education. Especially, in the era of digitalization and artificial intelligence, how to effectively cultivate skilled talents to achieve sustainable development of individuals and society has become a hot issue in continuing education. Considering the importance of this research topic, this paper introduces topic modeling methods including LDA and BERTopic with LLMs in a loop to comb through the past and present of lifelong learning research in CNKI and WOS, and propose a topic utility metric to gauge how effectively each discovered topic aligns with real educational needs. We found that CNKI research articles focus on eight topics: community education and learning organization construction, lifelong education for minorities, lifelong learning and modernization, information technology and online education, learning from foreign experience, platform construction, teacher improvement, credit bank construction, and global governance and multilateral cooperation. Research in WOS focuses on five aspects: policy support, professional personnel training, students’ self-learning, lifelong learning for the elderly, teacher training, and e-learning and digital literacy. Additionally, we observed that research in CNKI often has distinct Chinese characteristics, reflecting the country’s real conditions. For example, the “Internet + ” initiative features modern distance education, enabling students from remote areas to access courses offered in other cities. We encourage China to actively absorb international practical experience including providing students with a free and relaxed academic environment, cultivating students’ self-learning ability, and encouraging teachers to reflect on themselves for the purpose of having the high-quality development of lifelong education. Also, we suggest that future research incorporate competency-focused indicators into topic modeling to deepen understanding of the field’s development and enhance its policy relevance.