Meta-Analytical and Scientometric Review of Literature in Construction Engineering and Management

Abstract

This study executes a review of reviews utilizing a combinatorial mechanism of advanced text mining, unsupervised machine learning algorithms, and bibliometric network analyses deployed upon a corpus of 4,614 empirical review papers published between 1971 and 2026. Employing analytical Non-Negative Matrix Factorization procedures, the text corpus is stratified into eight distinct analytical thematic clusters. This multidimensional mapping reveals a transition diverging from historically entrenched legacy lifecycle management topologies and accelerating towards digitalization and machine learning integrations closely shadowed by smart self-healing material frameworks. The systemic analysis additionally employs a newly formulated Methodological Rigor Index to longitudinally track and prove the transition from opinion based narrative summaries into systematized evaluative reviews. Concurrent polynomial time series equations output empirical validation forecasting specific trajectory volumes through 2030. Furthermore, algorithmic calculation of lexical bursts and citation velocity quantifies the exact structural acceleration of specific conceptual frameworks, distinctly identifying white space research potentials.