Rider: Reader-Guided Passage Reranking for Open-Domain Question Answering
This work provides a simple and effective reranking solution for researchers and practitioners in open-domain question answering, offering significant performance gains without additional training.
This paper introduces RIDER, a training-free passage reranking method for open-domain question answering that uses a reader's top predictions to reorder retrieved passages. RIDER improves top-1 retrieval accuracy by 10-20 points and Exact Match (EM) by 1-4 points, achieving 48.3 EM on Natural Questions and 66.4 EM on TriviaQA with limited reader input.
Current open-domain question answering systems often follow a Retriever-Reader architecture, where the retriever first retrieves relevant passages and the reader then reads the retrieved passages to form an answer. In this paper, we propose a simple and effective passage reranking method, named Reader-guIDEd Reranker (RIDER), which does not involve training and reranks the retrieved passages solely based on the top predictions of the reader before reranking. We show that RIDER, despite its simplicity, achieves 10 to 20 absolute gains in top-1 retrieval accuracy and 1 to 4 Exact Match (EM) gains without refining the retriever or reader. In addition, RIDER, without any training, outperforms state-of-the-art transformer-based supervised rerankers. Remarkably, RIDER achieves 48.3 EM on the Natural Questions dataset and 66.4 EM on the TriviaQA dataset when only 1,024 tokens (7.8 passages on average) are used as the reader input after passage reranking.