A Comprehensive Review of Land Use and Land Cover Change Based on Knowledge Graph and Bibliometric Analyses

Abstract

Land use and land cover (LULC) changes are of vital significance in fields such as environmental impact assessment and natural disaster monitoring. This study, through an analysis of 1432 papers over the past decade employing quantitative, qualitative, bibliometric analysis, and knowledge graph techniques, aims to assess the evolution and current landscape of deep learning (DL) in LULC. The focus areas are: (1) trend analysis of the number and annual citations of published articles, (2) identification of leading institutions, countries/regions, and publication sources, (3) exploration of scientific collaborations among major institutions and countries/regions, and (4) examination of key research themes and their development trends. From 2013 to 2023 there was a substantial surge in the application of DL in LULC, with China standing out as the principal contributor. Notably, international cooperation, particularly between China and the USA, saw a significant increase. Furthermore, the study elucidates the challenges concerning sample data and models in the application of DL to LULC, providing insights that could guide future research directions to accelerate progress in this domain.