Paper Details

Abstract

Seed quality standards, as the minimum requirements for crop cultivation, guarantee the healthy growth of crops and are an essential basis for supervision by relevant departments. With the continuous development of China's standard system, the number of seed quality standard documents has increased dramatically, and the speed of updating has accelerated. The traditional paper, picture, or independent electronic document form is no longer adaptable to the current digital needs. It isn't easy to achieve rapid query and accurate positioning, and the relationship between the documents and content can not be embodied. To meet the internationalization requirements and make up for the above defects, this study first constructs the ontology architecture of seed quality standard documents based on the commonality of the structure and content of the standard document lines. Then regular expressions are performed for the structured data based on tables and semi-structured data based on text, and knowledge extraction is performed for the unstructured data by using the BERT-BiLSTM-CRF named entity recognition model. Finally, the graph is completed by using the Neo4j database to complete the construction of the graph, and achieve the information storage, visual display, multi-dimensional retrieval, and other functional applications. Seed quality standard knowledge mapping can more quickly and accurately obtain knowledge more intuitively understand the connection between various sub-quality standards, which can help non-researchers to locate and use swiftly and related personnel to conduct in-depth analysis and research, improve quality management efficiency and accuracy, and contribute digital power to plantation quality management.

To Cite This Article