Ontology-Based Decision Support Framework for Urban Road Renewal

Decision-making in Urban Road Renewal is often hindered by the disconnect between static conceptual models and dynamic industry specifications. To address this, this paper proposes an ontology-based decision support framework that formally models and integrates multi-source knowledge for automated compliance checking. A domain ontology was constructed by extracting entities from 15 key industry specifications using a BERT-BiLSTM-CRF deep learning model (achieving an accuracy of 98.2%), and a rule base of over 50 Semantic Web Rule Language (SWRL) rules was formulated to enable automated reasoning. The framework’s effectiveness was validated through a multi-agent simulation of the G15 Jialiu renewal project. Results demonstrated that the system-generated optimization measures increased traffic capacity by up to 95.0% and improved the Pavement Condition Index (PCI) by 6.1%, empirically verifying the directional consistency of the decision logic. Finally, the practical feasibility was demonstrated through a Decision Support System (DSS). This research provides a novel framework for leveraging fragmented knowledge, enhancing the consistency and rationality of decision-making in smart city infrastructure management.