Reasoning over Logically Interacted Conditions for Question Answering
This addresses a challenging reasoning task in NLP for question answering, but it is incremental as it builds on existing conditional QA datasets and methods.
The paper tackles the problem of question answering where answers depend on logically interacting conditions, some of which lack evidence, requiring models to perform logical reasoning and identify uncertain conditions. The proposed TReasoner model achieves state-of-the-art performance, outperforming previous methods by 3-10 points on benchmark datasets.
Some questions have multiple answers that are not equally correct, i.e. answers are different under different conditions. Conditions are used to distinguish answers as well as to provide additional information to support them. In this paper, we study a more challenging task where answers are constrained by a list of conditions that logically interact, which requires performing logical reasoning over the conditions to determine the correctness of the answers. Even more challenging, we only provide evidences for a subset of the conditions, so some questions may not have deterministic answers. In such cases, models are asked to find probable answers and identify conditions that need to be satisfied to make the answers correct. We propose a new model, TReasoner, for this challenging reasoning task. TReasoner consists of an entailment module, a reasoning module, and a generation module (if the answers are free-form text spans). TReasoner achieves state-of-the-art performance on two benchmark conditional QA datasets, outperforming the previous state-of-the-art by 3-10 points.