Missing Link
"Missing link" research encompasses diverse efforts to address incomplete or missing information across various domains. Current work focuses on predicting missing connections in graphs (using LLMs and graph neural networks), identifying relationships between labels in disparate datasets (leveraging visual and linguistic features), and improving the accuracy and efficiency of knowledge graph completion. These advancements have significant implications for improving automated decision-making systems, enhancing large language model capabilities, and facilitating more robust and explainable AI applications.
Papers
July 15, 2024
June 7, 2024
March 13, 2024
March 12, 2024
July 7, 2023
February 13, 2023
July 24, 2022