| Mangrove forests, as a vital component of coastal ecosystems, play a pivotal role in carbon sequestration, biodiversity conservation, and coastal protection. However, under the pressures of climate change and anthropogenic activities, these ecosystems are facing severe challenges, including health degradation, habitat fragmentation, and functional decline. To enable efficient organization and management of multi-source data related to mangrove habitat health and enhance scenario comprehension and data utility, we propose a knowledge graph construction method for multi-source heterogeneous data. First, a bottom-up approach was adopted to design semantic types and relationships for mangrove habitat health, establishing a unified knowledge framework. Next, the collected multi-source heterogeneous data, including remote sensing imagery, ecological monitoring records, meteorological and hydrological data, and literature, were subjected to knowledge extraction and processing using different strategies based on the constructed knowledge system. Finally, leveraging graph database technology, we implemented knowledge graph storage and demonstrated its application through information retrieval and visualization for mangrove habitat health assessment. The research results can provide a data foundation for studies on mangrove health state prediction and assessment based on knowledge graphs and deep learning, and offer a reference for related knowledge-graph research in the mangrove field. |
