From Examples to Rules: Neural Guided Rule Synthesis for Information Extraction
From Examples to Rules: Neural Guided Rule Synthesis for Information Extraction
复制标题
DOI：
--
复制
发表时间：
2022-01
期刊：
影响因子：
--
通讯作者：
Robert Vacareanu;M. A. Valenzuela-Escarcega;George C. G. Barbosa;Rebecca Sharp;M. Surdeanu
中科院分区：
文献类型：
--
作者：
Robert Vacareanu;M. A. Valenzuela-Escarcega;George C. G. Barbosa;Rebecca Sharp;M. Surdeanu
虽然信息提取的深度学习方法取得了许多成功，但随着需求的变化，它们可能难以增强或维护。另一方面，基于规则的方法可以更容易地修改。然而，制定规则需要语言学和感兴趣领域的专业知识，这对于大多数用户来说是不可行的。在这里，我们尝试结合这两个方向的优点，同时减轻它们的缺点。我们将程序综合相关领域的最新进展应用于信息提取，从提供的示例中综合规则。我们使用基于变压器的架构来指导枚举搜索，并表明这减少了找到规则之前需要探索的步骤数。此外，我们表明，无需在特定领域训练合成算法，我们的合成规则就可以在专注于关系分类的少样本学习的任务的 1-shot 场景中实现最先进的性能，并在 5-shot 场景中实现竞争性能。
While deep learning approaches to information extraction have had many successes, they can be difficult to augment or maintain as needs shift. Rule-based methods, on the other hand, can be more easily modified. However, crafting rules requires expertise in linguistics and the domain of interest, making it infeasible for most users. Here we attempt to combine the advantages of these two directions while mitigating their drawbacks. We adapt recent advances from the adjacent field of program synthesis to information extraction, synthesizing rules from provided examples. We use a transformer-based architecture to guide an enumerative search, and show that this reduces the number of steps that need to be explored before a rule is found. Further, we show that without training the synthesis algorithm on the specific domain, our synthesized rules achieve state-of-the-art performance on the 1-shot scenario of a task that focuses on few-shot learning for relation classification, and competitive performance in the 5-shot scenario.