Development and application of a field knowledge graph and search engine for pavement engineering

Abstract

Abstract Integrated, timely data about pavement structures, materials and performance information are crucial for the continuous improvement and optimization of pavement design by the engineering research community. However, at present, pavement structures, materials and performance information in China are relatively isolated and cannot be integrated and managed. This results in a waste of a large amount of effective information. One of the significant development trends of pavement engineering is to collect, analyze, and manage the knowledge assets of pavement information to realize intelligent decision-making. To address these challenges, a knowledge graph (KG) is adopted, which is a novel and effective knowledge management technology and provides an ideal technical method to realize the integration of information in pavement engineering. First, a neural network model is used based on the principle of deep learning to obtain knowledge. On this basis, the relationship between knowledge is built from siloed databases, data in textual format and networks, and the knowledge base. Second, KG-Pavement is presented, which is a flexible framework that can integrate and ingest heterogeneous pavement engineering data to generate knowledge graphs. Furthermore, the index and unique constraints on attributes for knowledge entities are proposed in KG-Pavement, which can improve the efficiency of internal retrieval in the system. Finally, a pavement information search engine based on a knowledge graph is constructed to realize information interaction and target information matching between a webpage server and graph database. This is the first successful application of knowledge graphs in pavement engineering. This will greatly promote knowledge integration and intelligent decision-making in the domain of pavement engineering.