How Well Do Multi-hop Reading Comprehension Models Understand Date Information?
This work addresses a specific limitation in multi-hop QA systems for researchers and practitioners, offering a dataset and evaluation method to enhance model robustness, though it is incremental as it builds on existing datasets.
The authors tackled the problem of evaluating multi-hop reading comprehension models' ability to perform step-by-step reasoning on date information, revealing that these models fail at date subtraction despite performing well in related tasks, and showing that their proposed probing questions can improve model performance by up to +10.3 F1.
Several multi-hop reading comprehension datasets have been proposed to resolve the issue of reasoning shortcuts by which questions can be answered without performing multi-hop reasoning. However, the ability of multi-hop models to perform step-by-step reasoning when finding an answer to a comparison question remains unclear. It is also unclear how questions about the internal reasoning process are useful for training and evaluating question-answering (QA) systems. To evaluate the model precisely in a hierarchical manner, we first propose a dataset, \textit{HieraDate}, with three probing tasks in addition to the main question: extraction, reasoning, and robustness. Our dataset is created by enhancing two previous multi-hop datasets, HotpotQA and 2WikiMultiHopQA, focusing on multi-hop questions on date information that involve both comparison and numerical reasoning. We then evaluate the ability of existing models to understand date information. Our experimental results reveal that the multi-hop models do not have the ability to subtract two dates even when they perform well in date comparison and number subtraction tasks. Other results reveal that our probing questions can help to improve the performance of the models (e.g., by +10.3 F1) on the main QA task and our dataset can be used for data augmentation to improve the robustness of the models.