Coarse-grained decomposition and fine-grained interaction for multi-hop question answering
It addresses the problem of complex queries requiring multi-hop reasoning for question answering systems, with incremental improvements over existing methods.
The paper tackles multi-hop question answering by proposing a model that decomposes complex questions into simpler ones and enhances semantic interactions, achieving state-of-the-art performance on SQuAD and HotpotQA datasets.
Recent advances regarding question answering and reading comprehension have resulted in models that surpass human performance when the answer is contained in a single, continuous passage of text, requiring only single-hop reasoning. However, in actual scenarios, lots of complex queries require multi-hop reasoning. The key to the Question Answering task is semantic feature interaction between documents and questions, which is widely processed by Bi-directional Attention Flow (Bi-DAF), but Bi-DAF generally captures only the surface semantics of words in complex questions and fails to capture implied semantic feature of intermediate answers. As a result, Bi-DAF partially ignores part of the contexts related to the question and cannot extract the most important parts of multiple documents. In this paper we propose a new model architecture for multi-hop question answering, by applying two completion strategies: (1) Coarse-Grain complex question Decomposition (CGDe) strategy are introduced to decompose complex question into simple ones under the condition of without any additional annotations (2) Fine-Grained Interaction (FGIn) strategy are introduced to better represent each word in the document and extract more comprehensive and accurate sentences related to the inference path. The above two strategies are combined and tested on the SQuAD and HotpotQA datasets, and the experimental results show that our method outperforms state-of-the-art baselines.