Identifying Circular City Indicators Based on Advanced Text Analytics: A Multi-Algorithmic Approach

Circular Economy (CE) and circular cities are recognized as essential approaches for achieving sustainability and fostering sustainable urban development. Given the diverse definitions and principles, multidimensional complexities, and lack of a comprehensive list of CE indicators, this study aims to propose an innovative method for identifying macro-level indicators to assess urban circularity. This methodology combines a systematic literature review (SLR) with advanced machine learning (ML) and natural language processing (NLP) techniques. A multi-algorithmic approach, incorporating BERT, TF-IDF, Word2Vec, graph-based and clustering models, is employed to extract a comprehensive set of indicators from reputable scientific articles and reports to compare frequency and similarly based on each model. The overlap and accuracy of results from these five methods are analyzed to produce a refined list of indicators with high precision and alignment with core CE principles. This curated collection serves as a valuable tool for policymakers, urban planners, and designers, enabling the prediction of future trends in urban circularity. Additionally, it provides guidance for research and practical projects at various scales, from buildings and neighborhoods to entire cities, facilitating a more precise assessment of sustainability and circularity in modern urban environments.