Аңдатпа
Despite the urgency to transition to a Circular Economy (CE) in construction, circularity principles remain inconsistently implemented and difficult to monitor across stakeholders and regulatory contexts. This study employs Natural Language Processing (NLP) to examine how circular construction is defined, debated, and operationalised across multiple levels and discourses. A multi-level analysis (academic, social media, regulatory levels) is introduced to trace how circularity concepts travel from theory to practice. At the academic level, NLP analysis of 480 articles reveals a conceptual divergence between circularity objectives focused on operational resource efficiency and waste management, and broader sustainability objectives addressing long-term environmental impacts. This conceptual ambiguity can hinder coherent implementation. At the public-discourse level, sentiment analysis shows that 57% of stake-holder opinions on construction circularity are positive, highlighting which concepts resonate and which remain abstract. At the policy level, analysis of CE strategies from 57 countries identifies five implementation clusters based on the 10R hierarchy, demonstrating that adoption patterns are shaped more by institutional capacity than economic development. At the technical level, text-mining of standards in Central Asia shows that many CE principles promoted in academia and policy fail to materialise in operational requirements. This multi-level approach reveals three critical dynamics: (1) vertical disconnects between policy ambitions and technical standards; (2) horizontal variation in national interpretations of circularity; and (3) missing feedback loops between implementation challenges and strategic or research agendas. NLP provides a diagnostic tool enabling stakeholders to identify misalignments and understand how circularity concepts diffuse—or stall—across the construction ecosystem.