Mapping the Research Agenda for Sustainable Fashion Design: Constructing Knowledge Maps Through Bibliometric Analysis and LDA Machine Learning

This study aims to construct dynamic knowledge maps and uncover the thematic evolution of research on fashion design for sustainable development (FDSD), using bibliometric analysis and topic modeling to identify knowledge structures and emerging trends. As global sustainability goals exert increasing pressure on the fashion industry, FDSD has emerged as a critical field in both academic and practical domains. Mapping its intellectual development is essential to support informed scholarly inquiry and guide sustainable innovation. Based on 879 peer-reviewed articles retrieved from the Web of Science Core Collection (2004–2024), we employed a mixed-methods approach. CiteSpace was used to visualize collaborative structures across authors, institutions, and countries, as well as keyword co-occurrence patterns. Latent Dirichlet Allocation (LDA) modeling was applied to extract latent topics and trace their temporal shifts. The results show a notable rise in publications since 2020, with major contributions from Federico Caniato and Tsan-Ming Choi, and leading institutions including Hong Kong Polytechnic University and Politecnico di Milano. High-frequency keywords such as “sustainability,†“circular economy,†and “corporate social responsibility†reveal evolving research priorities. LDA analysis illustrates a thematic shift from eco-materials and life-cycle analysis to ethical consumption, supply chain transparency, and circular business models. This study provides a comprehensive overview of the FDSD research landscape. The findings offer valuable insights for scholars refining theoretical frameworks, educators aligning curricula with emerging themes, and industry practitioners and policymakers integrating sustainability into fashion design and governance.