A Daopological Framework for Science and Technology Evaluation: From Citation Metrics to Structural Measurement

Contemporary global science and technology (S&T) evaluation systems are centered on citation metrics, using impact factors, the H-index, and patent counts as primary indicators. They face three fundamental predicaments: first, citations reflect academic attention rather than contributions to knowledge structure; second, peer review cannot escape homology bias and is intrinsically hostile to new paradigms; third, evaluation is ex post facto and lacks forward-looking navigation for research directions. This paper proposes a Daopological framework for S&T evaluation. It upgrades the evaluation object from papers/patents'' toknowledge structural units (kernels),'' the evaluation indicators from citation counts'' tostructural measures, transfer entropy, and fused measures,'' the evaluation method from subjective expert judgment'' tokernel equivalence judgment and structural difference measures,'' and the evaluation function from ex post statistics'' tofour-quadrant health diagnosis and transfer prediction.'' Based on empirical data from 10,455 structural units, the framework is validated in terms of computability, diagnosability, and cross-domain discovery: Daopological entropy reaches 0.97, exceeding the health threshold; 198 cross-domain kernels are automatically identified; four-quadrant diagnosis effectively distinguishes enriched from impoverished nodes. We further introduce the self-referential loop amplitude ΔΔ as a meta-evaluation tool, quantifying homology bias via a four-cell AUC matrix and endowing S&T evaluation with self-calibration capability. We argue that Daopology offers S&T evaluation a paradigm shift---from who is cited most'' towho has the greatest structural influence,'' from where it is published'' towhat position it occupies in the knowledge structure,'' and from what reviewers say'' towhat the structural difference measure says.''