Construction Of Knowledge Graph Based On Soybean Breeding Corpus

Soybean is an important source of protein and edible oil.At present,domestic soybean is heavily dependent on import,which has threatened the national grain and oil security.In order to change the current situation,it is urgent for domestic soybean breeding experts to change the traditional breeding methods and cultivate new high-yield soybean varieties as soon as possible.Based on the knowledge graph technology,this paper presents the numerous historical soybean breeding knowledge in the way to understand easily,and provides big data support for breeding experts to design accurate breeding schemes.This thesis mainly completed the following work:（1）Taking the Internet as one of the information sources of breeding knowledge and introducing natural language processing technologies such as named entity recognition and relationship extraction,a complete set of soybean breeding knowledge graph information extraction scheme is designed to extract soybean breeding information from Internet data timely and accurately.（2）For the soybean breeding knowledge records accumulated by each domestic breeding team over the years,after preprocessing and integration,the three tables were integrated into four relational databases and the genetic relationships were extracted when different varieties of soybean seeds participated as breeding material.The affinity between soybeans was obtained through knowledge reasoning,and the knowledge graph of soybean breeding race lines was constructed.The breeding family tree is presented intuitively.（3）Take the existing breeding information system as one of the information sources,integrate the soybean breeding information in the network and the soybean breeding information provided by the Research Institute,complete the processing of soybean seed information,geographic environment information of experimental field and relationship information between soybean traits and environment,establish the relationship between the three,and further infer the environment suitable for the growth of different varieties of soybean.It can effetely guide the seed selection and sales.This paper is an exploration of introducing knowledge graph technology into the field of soybean breeding design.Through knowledge graph technology,the information integration of multiple data sources and the rapid seed test of soybean seeds are realized,which provides a reference for breeding experts to design soybean breeding scheme efficiently.