Leveraging Semantic and Lexical Matching to Improve the Recall of Document Retrieval Systems: A Hybrid Approach
Topics: AI (Deep Learning), Entity based search, Marc Najork, Ranking, Reranking, Semantic Search
The paper proposes a hybrid document retrieval approach that combines deep neural network models with traditional lexical models. This approach aims to enhance the initial retrieval stage, which is usually performed with models like BM25, by integrating semantic understanding through deep learning techniques.
This work provides valuable insights into improving document retrieval systems and underscores the potential of combining semantic and lexical models to enhance efficiency and effectiveness in the retrieval stage.