AUTHOR=Wang Kun , Miao Yuyuan , Wang Xu , Li Yuze , Li Fuzhong , Song Haiyan 

TITLE=Research on the construction of a knowledge graph for tomato leaf pests and diseases based on the named entity recognition model

JOURNAL=Frontiers in Plant Science

VOLUME=Volume 15 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2024.1482275

DOI=10.3389/fpls.2024.1482275

ISSN=1664-462X

ABSTRACT=Tomato leaf pests and diseases pose a serious threat to the yield and quality of tomatoes, making it crucial to deeply study control methods. However, current control measures mainly rely on experience and manual observation, which makes it difficult to fully integrate multi-source data. The lack of timely data and intelligent analysis support significantly undermines the scientific accuracy of control decisions, making it challenging to effectively address the complex and dynamic issues of pests and diseases. To address this, we integrated tomato leaf pest and disease information resources, gathering data from agricultural standards documents, agricultural knowledge websites, and literature. Under the guidance of domain experts, we preprocessed the data to construct a sample set of tomato leaf pests and diseases. We then utilized the Named Entity Recognition (NER) model ALBERT-BiLSTM-CRF to conduct end-to-end knowledge extraction experiments, which showed better results than classic models such as 1DCNN-CRF and BiLSTM-CRF, with a recall rate of 95.03%. Finally, the extracted knowledge was stored in the Neo4j graph database, visually reflecting the internal structure of the knowledge graph. We developed a digital diagnostic system for tomato leaf pests and diseases based on the knowledge graph, achieving graphical management and visualization of tomato leaf pest and disease knowledge. The knowledge graph constructed in this study can provide suggestions for tomato leaf pest and disease control and offer new research ideas for pest control of other crops.