Phrase Retrieval
Phrase retrieval focuses on efficiently finding specific phrases within large text corpora, aiming to improve information retrieval beyond keyword matching. Current research emphasizes developing robust phrase representations using contextualized embeddings and contrastive learning, often incorporating graph-based methods to capture broader semantic relationships or handling complex, multi-faceted queries. These advancements are significantly impacting downstream NLP tasks like question answering and machine translation, improving accuracy and efficiency by leveraging the fine-grained information provided by phrases.
Papers
November 16, 2024
May 12, 2024
March 25, 2024
March 24, 2024
June 7, 2023
May 24, 2023