Knowledge Engineering Standards Committee

Active Standards and Current Projects

IEEE P2807.5 Guide for Medical Clinical Diagnosis and Treatment Oriented Knowledge Graphs

IEEE P2807.8 Standard for knowledge exchange and fusion protocol among knowledge graphs

IEEE P2807.10 Guide for Knowledge Modeling for Carbon Verification-Oriented Knowledge Graphs

IEEE P2807.12 Standard for technical requirements for general knowledge services based on knowledge graphs

IEEE P2807.13 Guide for integration framework among large-scale pretrained model and knowledge graphs

Knowledge graphs have been widely applied into smart finance, smart grid, smart healthcare, smart education and other domains. Knowledge services based on knowledge graphs meet a critical demand to many enterprises and organizations today by providing the factual knowledge and discovering potential knowledge which they are interested in or need. These standards aim to address the customer need for a specific and objective way to construct the knowledge graphs and evaluate the performance of general knowledge services based on knowledge graphs, including intelligent retrieval, intelligent recommendation, assisted decision-making, knowledge question and answer, etc.

In addition, large-scale pre-trained models can generally provide stronger generalizability, less sensitiveness to model structures and lower test errors, in comparison with small-scale models. However, the knowledge graphs face some difficulties, such as long construction process, high cost on ontology designing and knowledge acquisition from unstructured data, lack of common sense knowledge; the large-scale pre-trained models also face some difficulties, such as lack of domain knowledge, insufficient update frequency of knowledge, insufficient visualization of large-scale knowledge and their relationship. There is strong complementarity among knowledge graphs and large-scale pre-trained models. Integration among them can generate more value for the users. Thus, the standards also aim to address the need for the collaboration of knowledge graph providers, integrators and model providers.

All types and sizes of organizations, including public and private companies, government entities, and not-for-profit organizations, that are designing, implementing or using knowledge graph systems and large-scale pre-trained models.

IEEE P2807.5 Guide for Medical Clinical Diagnosis and Treatment Oriented Knowledge Graphs

IEEE P2807.8 Standard for knowledge exchange and fusion protocol among knowledge graphs

IEEE P2807.10 Guide for Knowledge Modeling for Carbon Verification-Oriented Knowledge Graphs

IEEE P2807.12 Standard for technical requirements for general knowledge services based on knowledge graphs

IEEE P2807.13 Guide for integration framework among large-scale pretrained model and knowledge graphs

Knowledge graphs have been widely applied into smart finance, smart grid, smart healthcare, smart education and other domains. Knowledge services based on knowledge graphs meet a critical demand to many enterprises and organizations today by providing the factual knowledge and discovering potential knowledge which they are interested in or need. These standards aim to address the customer need for a specific and objective way to construct the knowledge graphs and evaluate the performance of general knowledge services based on knowledge graphs, including intelligent retrieval, intelligent recommendation, assisted decision-making, knowledge question and answer, etc.

In addition, large-scale pre-trained models can generally provide stronger generalizability, less sensitiveness to model structures and lower test errors, in comparison with small-scale models. However, the knowledge graphs face some difficulties, such as long construction process, high cost on ontology designing and knowledge acquisition from unstructured data, lack of common sense knowledge; the large-scale pre-trained models also face some difficulties, such as lack of domain knowledge, insufficient update frequency of knowledge, insufficient visualization of large-scale knowledge and their relationship. There is strong complementarity among knowledge graphs and large-scale pre-trained models. Integration among them can generate more value for the users. Thus, the standards also aim to address the need for the collaboration of knowledge graph providers, integrators and model providers.

All types and sizes of organizations, including public and private companies, government entities, and not-for-profit organizations, that are designing, implementing or using knowledge graph systems and large-scale pre-trained models.