Tracing the evolution of household carbon emission research by machine learning

As a major driver of global production, household consumption warrants increased attention due to its potential for decarbonization. This study systematically reviews 4647 journal papers to uncover evolving research themes and regional disparities in household carbon emissions research. We identified a marked increase in publications after 2015 and, using natural language processing, defined 7 distinct research clusters. While earlier studies focused primarily on energy consumption, more recent research has broadened to encompass a range of topics, including food systems, supply-chain emissions, and renewable energy. Our regional analysis reveals varied levels of research engagement across countries. The Global South is increasingly involved, focusing on cooking facilities and household energy use, whereas Europe emphasizes policy-oriented mitigation. North America, led by the United States, focuses on transportation emissions and residential energy-efficiency modeling, while Asian countries examine urban-rural disparities and household inequality. Drawing on these thematic and geographical insights, this study identifies current knowledge gaps, such as the need for cross-regional comparisons and supply chain integration. These insights offer actionable directions to promote sustainable consumption practices and support global climate goals. • Apply machine learning to review 4647 papers on household carbon emissions. • Identify shifts across 7 thematic categories in household carbon research. • Highlight emerging fields, including renewable energy and food systems • Reveal regional disparities in research focus across continents. • Unveil future research directions for household decarbonization strategies.