Contaminated sites are a main cause of urban soil problems and have led to increasing pollution and public risk in China as a result of the rapid growth of industrial and urban land use. Because land pollution involves extensive multi-source heterogeneous information, identifying the risk of urban soil pollution efficiently and predicting pollution-related events are important for urban environmental management. Knowledge graphs (KGs) have unique advantages in dealing with massive amounts of information. This study attempts to construct a KG of contaminated sites in South China to explore its feasibility and effectiveness in urban soil environmental management. The results demonstrate that KGs have a favorable effect in information retrieval, knowledge reasoning, and visualization. Studied cases in this article demonstrate that the KG model can achieve many functions, including the display of global information of polluted sites, and discovery of regional distribution of characteristic pollutants and main pollutants of specific industries, based on special query syntax. However, this approach is limited by some technical difficulties, such as knowledge mining of natural resources, which must be overcome in future studies to improve the operability of KG technologies.
Keywords: Contaminated site; Environmental management; Knowledge graph; Urban soil pollution.
Copyright © 2022 The Authors. Published by Elsevier Ltd.. All rights reserved.