Natural Language Deduction through Search over Statement Compositions
Natural Language Deduction through Search over Statement Compositions
复制标题
DOI：
10.18653/v1/2022.findings-emnlp.358
复制
发表时间：
2022-01
期刊：
影响因子：
--
通讯作者：
Kaj Bostrom;Zayne Sprague;Swarat Chaudhuri;Greg Durrett
中科院分区：
文献类型：
--
作者：
Kaj Bostrom;Zayne Sprague;Swarat Chaudhuri;Greg Durrett
在从事实核查到问题回答的环境中，我们经常想知道收集的证据(前提)是否需要一个假设。现有的方法主要集中在这一任务的端到端区分版本，但很少有工作处理生成版本，在生成版本中，模型搜索前提所需的陈述空间以建设性地推导假设。我们提出了一个用自然语言进行这种演绎推理的系统，通过将任务分解成由搜索过程协调的单独步骤，产生忠实地反映系统推理过程的中间结论树。我们在EntramentBank数据集(Dalvi等人，2021年)上的实验表明，该系统可以成功地证明真实的陈述，同时拒绝错误的陈述。此外，它产生的自然语言解释的步骤效度比端到端T5模型的绝对效度高17%。
In settings from fact-checking to question answering, we frequently want to know whether a collection of evidence (premises) entails a hypothesis. Existing methods primarily focus on the end-to-end discriminative version of this task, but less work has treated the generative version in which a model searches over the space of statements entailed by the premises to constructively derive the hypothesis. We propose a system for doing this kind of deductive reasoning in natural language by decomposing the task into separate steps coordinated by a search procedure, producing a tree of intermediate conclusions that faithfully reflects the system's reasoning process. Our experiments on the EntailmentBank dataset (Dalvi et al., 2021) demonstrate that the proposed system can successfully prove true statements while rejecting false ones. Moreover, it produces natural language explanations with a 17% absolute higher step validity than those produced by an end-to-end T5 model.