Document Relevance
Document relevance research focuses on efficiently identifying and utilizing information within large document collections to answer queries accurately. Current efforts concentrate on improving retrieval methods, particularly through large language models (LLMs) and graph-based approaches, and enhancing the effectiveness of retrieval-augmented generation (RAG) systems by addressing challenges like context compression and handling unanswerable questions. These advancements are crucial for improving information access in various applications, including search engines, eDiscovery, and clinical trial document generation, ultimately impacting efficiency and accuracy in knowledge-intensive tasks.
Papers
DSI++: Updating Transformer Memory with New Documents
Sanket Vaibhav Mehta, Jai Gupta, Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Jinfeng Rao, Marc Najork, Emma Strubell, Donald Metzler
Human in the loop: How to effectively create coherent topics by manually labeling only a few documents per class
Anton Thielmann, Christoph Weisser, Benjamin Säfken