OpenQA System
Open-domain question answering (OpenQA) systems aim to answer factual questions using large-scale knowledge sources, addressing the challenge of retrieving and processing information from diverse and evolving data. Current research focuses on improving model generalization across different knowledge domains and languages, often employing retrieval-augmented architectures that combine information retrieval techniques with large language models, and exploring methods to mitigate over-reliance on memorized training data. These advancements are significant for improving access to information, particularly in low-resource languages, and for developing more robust and reliable question-answering systems across various applications.
Papers
April 2, 2024
March 1, 2024
January 7, 2024
November 6, 2023
May 23, 2023
October 26, 2022
October 11, 2022
April 1, 2022