论文标题
轻巧的CNN的AMR质量评级
AMR Quality Rating with a Lightweight CNN
论文作者
论文摘要
结构化的语义句子表示，例如抽象含义表示（AMR）在各种NLP任务中可能有用。但是，自动解析的质量可能会有很大的变化，并危害它们的用处。在没有昂贵的黄金数据的情况下，可以准确评估AMR质量的模型可以减轻这种情况，从而使我们能够将Incorporated Parse的可信赖性或在不同候选菜单中进行选择。 在这项工作中，我们建议将AMR图传输到图像域。这使我们能够创建一个简单的卷积神经网络（CNN），该网络模仿了人类法官，该法官负责评级图质量。我们的实验表明，该方法在几个质量维度上可以比强基础更准确地对质量进行评分。此外，该方法被证明是有效的，并降低了所产生的能耗。
Structured semantic sentence representations such as Abstract Meaning Representations (AMRs) are potentially useful in various NLP tasks. However, the quality of automatic parses can vary greatly and jeopardizes their usefulness. This can be mitigated by models that can accurately rate AMR quality in the absence of costly gold data, allowing us to inform downstream systems about an incorporated parse's trustworthiness or select among different candidate parses. In this work, we propose to transfer the AMR graph to the domain of images. This allows us to create a simple convolutional neural network (CNN) that imitates a human judge tasked with rating graph quality. Our experiments show that the method can rate quality more accurately than strong baselines, in several quality dimensions. Moreover, the method proves to be efficient and reduces the incurred energy consumption.
