Neural Network Acceptability Judgments
Neural Network Acceptability Judgments
复制标题
DOI：
10.1162/tacl_a_00290
复制
发表时间：
2019-01-01
影响因子：
10.9
通讯作者：
Bowman, Samuel R.
中科院分区：
文献类型：
--
作者：
Warstadt, Alex;Singh, Amanpreet;Bowman, Samuel R.
本文研究了人工神经网络判断句子的语法可接受性的能力，目的是测试其语言能力。我们介绍了语言可接受性（COLA）的语料库，这是一组10,657个英语句子，标记为来自发表的语言学文献中的语法或语法。作为基础线，我们训练了几种经常性神经网络模型的可接受性分类，并发现我们的模型优于Lau等人的无监督模型。 （2016年）在可乐。对特定语法现象的误差分析表明，Lau等人的模型和我们的模型都学到了系统的概括，例如主题 - 动物对象顺序。但是，我们测试的所有模型在广泛的语法结构上的性能远低于人类水平。
This paper investigates the ability of artificial neural networks to judge the grammatical acceptability of a sentence, with the goal of testing their linguistic competence. We introduce the Corpus of Linguistic Acceptability (CoLA), a set of 10,657 English sentences labeled as grammatical or ungrammatical from published linguistics literature. As baselines, we train several recurrent neural network models on acceptability classification, and find that our models outperform unsupervised models by Lau et al. (2016) on CoLA. Error-analysis on specific grammatical phenomena reveals that both Lau et al.'s models and ours learn systematic generalizations like subject-verb-object order. However, all models we test perform far below human level on a wide range of grammatical constructions.