RocketQA: An Optimized Training Approach to Dense Passage Retrieval for Open-Domain Question Answering
This work addresses the problem of improving retrieval accuracy for open-domain question answering systems, representing an incremental advancement with specific technical contributions.
The paper tackles the challenges of training dual-encoders for dense passage retrieval in open-domain question answering by proposing RocketQA, an optimized training approach using cross-batch negatives, denoised hard negatives, and data augmentation, which significantly outperforms previous state-of-the-art models on MSMARCO and Natural Questions datasets.
In open-domain question answering, dense passage retrieval has become a new paradigm to retrieve relevant passages for finding answers. Typically, the dual-encoder architecture is adopted to learn dense representations of questions and passages for semantic matching. However, it is difficult to effectively train a dual-encoder due to the challenges including the discrepancy between training and inference, the existence of unlabeled positives and limited training data. To address these challenges, we propose an optimized training approach, called RocketQA, to improving dense passage retrieval. We make three major technical contributions in RocketQA, namely cross-batch negatives, denoised hard negatives and data augmentation. The experiment results show that RocketQA significantly outperforms previous state-of-the-art models on both MSMARCO and Natural Questions. We also conduct extensive experiments to examine the effectiveness of the three strategies in RocketQA. Besides, we demonstrate that the performance of end-to-end QA can be improved based on our RocketQA retriever.