The article discusses the challenges in Knowledge Graph (KG) learning due to its immense size. Four key deficiencies are identified: lack of expert knowledge integration, instability to node degree extremity in the KG, lack of consideration for uncertainty and relevance while learning, and lack of explainability. The authors propose a holistic approach to overcome these challenges, emphasizing that addressing these issues individually rather than as a whole is hindering human-KG empowerment. They present the ‘Veni, Vidi, Vici’ framework as a roadmap for shifting to a holistic co-empowerment model in KG learning and the broader machine learning domain.
Publication date: 12 Feb 2024
Project Page: Not provided
Paper: https://arxiv.org/pdf/2402.06098
