Context-Aware Answer Extraction in Question Answering
This addresses a specific issue in question answering for improving model reliability in real-world applications, representing an incremental advance.
The paper tackled the problem of extractive QA models predicting correct answer text but in irrelevant contexts, especially when answer text appears multiple times in a passage, by proposing BLANC with context prediction and block attention, resulting in outperforming state-of-the-art QA models with increasing performance gaps as answer occurrences rise and achieving better zero-shot results.
Extractive QA models have shown very promising performance in predicting the correct answer to a question for a given passage. However, they sometimes result in predicting the correct answer text but in a context irrelevant to the given question. This discrepancy becomes especially important as the number of occurrences of the answer text in a passage increases. To resolve this issue, we propose \textbf{BLANC} (\textbf{BL}ock \textbf{A}ttentio\textbf{N} for \textbf{C}ontext prediction) based on two main ideas: context prediction as an auxiliary task in multi-task learning manner, and a block attention method that learns the context prediction task. With experiments on reading comprehension, we show that BLANC outperforms the state-of-the-art QA models, and the performance gap increases as the number of answer text occurrences increases. We also conduct an experiment of training the models using SQuAD and predicting the supporting facts on HotpotQA and show that BLANC outperforms all baseline models in this zero-shot setting.