Contextual Semantic Parsing for Multilingual Task-Oriented Dialogues
Contextual Semantic Parsing for Multilingual Task-Oriented Dialogues
复制标题
DOI：
10.18653/v1/2023.eacl-main.63
复制
发表时间：
2021-11
期刊：
影响因子：
--
通讯作者：
M. Moradshahi;Victoria Tsai;Giovanni Campagna;M. Lam
中科院分区：
文献类型：
--
作者：
M. Moradshahi;Victoria Tsai;Giovanni Campagna;M. Lam
面向任务的对话系统的鲁棒状态跟踪目前仍然仅限于几种流行的语言。本文表明，给定一种语言的大规模对话数据集，我们可以使用机器翻译自动为其他语言生成有效的语义解析器。我们建议自动翻译对话数据集并进行对齐，以确保槽值的忠实翻译，并消除先前基准测试中使用的昂贵的人工监督。我们还提出了一种新的上下文语义解析模型，该模型对形式槽和值进行编码，并且仅对最后的代理和用户话语进行编码。我们表明，简洁的表示减少了翻译错误的复合效应，而不会损害实践中的准确性。我们在几个对话状态跟踪基准上评估了我们的方法。在 RiSAWOZ、CrossWOZ、CrossWOZ-EN 和 MultiWOZ-ZH 数据集上，我们将联合目标准确度提高了 11%、17%、20% 和 0.3%。我们对所有三个数据集进行了全面的错误分析，显示错误的注释可能会导致对模型质量的错误判断。最后，我们展示使用我们的翻译方法创建的 RiSAWOZ 英语和德语数据集。在这些数据集上，准确度在原始数据的 11% 以内，这表明高精度多语言对话数据集无需依赖昂贵的人工注释即可实现。我们开源数据集和软件。
Robust state tracking for task-oriented dialogue systems currently remains restricted to a few popular languages.This paper shows that given a large-scale dialogue data set in one language, we can automatically produce an effective semantic parser for other languages using machine translation. We propose automatic translation of dialogue datasets with alignment to ensure faithful translation of slot values and eliminate costly human supervision used in previous benchmarks. We also propose a new contextual semantic parsing model, which encodes the formal slots and values, and only the last agent and user utterances. We show that the succinct representation reduces the compounding effect of translation errors, without harming the accuracy in practice.We evaluate our approach on several dialogue state tracking benchmarks. On RiSAWOZ, CrossWOZ, CrossWOZ-EN, and MultiWOZ-ZH datasets we improve the state of the art by 11%, 17%, 20%, and 0.3% in joint goal accuracy. We present a comprehensive error analysis for all three datasets showing erroneous annotations can lead to misguided judgments on the quality of the model. Finally, we present RiSAWOZ English and German datasets, created using our translation methodology. On these datasets, accuracy is within 11% of the original showing that high-accuracy multilingual dialogue datasets are possible without relying on expensive human annotations. We release our datasets and software open source.