Scientific Knowledge Dependency Infrastructure (SKDI): An Open, Version-Controlled Framework for the Traceability, Evolution and Systematic Reassessment of Scientific Knowledge

Scientific Knowledge Dependency Infrastructure (SKDI) proposes an open, interoperable and version-controlled framework for representing the dependencies underlying scientific knowledge. The framework investigates how foundational assumptions, observations, measurements, experimental methodologies, analytical procedures, theoretical models and scientific conclusions can be represented as interconnected, independently identifiable and versioned components. Inspired by distributed version-control principles, SKDI aims to enable historical reconstruction of scientific knowledge and systematic identification of potentially affected downstream conclusions when foundational assumptions or methodologies change. The proposal builds upon established research in scientific provenance, semantic knowledge representation, evidence ontologies, scientific knowledge graphs and research object management. Its central research objective is to investigate the integration of these capabilities into an infrastructure supporting the traceability, evolution and reassessment of scientific knowledge. SKDI does not seek to automate scientific judgment or determine scientific truth. It aims to provide researchers and institutions with tools for examining the foundations and dependencies underlying scientific conclusions. Status: Foundational conceptual proposal, Version 1.0. Technical architecture, reference implementation, interoperability specifications and experimental validation are proposed areas of subsequent development.