Abstract: [Purpose/Significance] In recent years, challenges such as pandemics, wars, and natural disasters have posed numerous threats to China's food security. As the core of future agricultural productivity improvement, the importance of the seed industry has been continuously emphasized by the government. To facilitate preservation and utilization, scholars have integrated and digitized a vast amount of germplasm resources. However, the current platforms of crop germplasm resource knowledge services still suffer from issues such as diverse and fragmented large-scale heterogeneous data sources, lack of interconnection among data, and insufficient exploration of the data, thereby falling short of achieving intelligent and semantic research on germplasm resources. Therefore, this article aims to propose an effective method for knowledge organization and semantic association to meet the growing demand for intelligent knowledge services from users. The proposed method is to provide insights into the development of germplasm resource knowledge service platforms tailored for computational breeding. [Method/Process] This paper conducted a comparative analysis by examining the description and organization of germplasm resource data domestically and internationally. Four mainstream international platforms of germplasm resource knowledge services were selected for comparison from five perspectives: general overview, resource quantity, the types of knowledge, retrieval methods, and results. The deficiencies of these platforms in intelligent services such as text mining, semantic retrieval, and knowledge computation were summarized. In general, these platforms still rely on keyword-based retrieval as the primary means of searching, lacking systematic modeling of germplasm resource knowledge and the ability to achieve semantic retrieval in an intelligent environment. However, with the development of information technology in the era of big data, there is a growing demand in China to promote the development of computational breeding and provide more accurate, faster, and more intelligent knowledge resources to researchers and ordinary farmers through AI-based germplasm resource knowledge services. Therefore, in response to these new demands, the article proposes the construction of a panoramic crop germplasm resource knowledge graph and the development of a knowledge graph-driven germplasm resource knowledge service platform. [Results/Conclusions] The knowledge graph provides a more efficient and intelligent form of knowledge organization, and a knowledge service platform based on the knowledge graph contributes to improved efficiency and accuracy of knowledge services. In the next step, this research will focus on building a large-scale germplasm resource knowledge graph based on germplasm resource data and expanding it with other data, such as genotype, phenotype, environmental, and literature information. The application exploration will be conducted in scenarios such as intelligent question answering and knowledge-based computational breeding.

Key words: crops, germplasm resources, computational breeding, knowledge service, knowledge graph