|
|||||
|
|
| 基于BERT的意图分类与槽填充联合方法 | |
| 引用本文： | 马天宇,覃俊,刘晶,帖军,后琦.基于BERT的意图分类与槽填充联合方法[J].中文信息学报,2022,36(8):127-134. |
| 作者姓名： | 马天宇 覃俊 刘晶 帖军 后琦 |
| 作者单位： | 1.中南民族大学 计算机科学学院,湖北 武汉 430074; 2.湖北省制造企业智能管理工程技术研究中心,湖北 武汉 430074 |
| 基金项目： | 湖北省技术创新专项重大项目(2019ABA101);武汉市科技计划应用基础前沿项目(2020020601012267);中南民族大学中央高校基本科研业务费专项(CZQ20012) |
| 摘 要： | 口语理解是自然语言处理的一个重要内容,意图分类和槽填充是口语理解的两个基本子任务。最近的研究表明,共同学习这两项任务可以起到相互促进的作用。该文提出了一个基于BERT的意图分类联合模型,通过一个关联网络使得两个任务建立直接联系和共享信息,以此来提升任务效果。模型引入BERT来增强词向量的语义表示,有效解决了目前联合模型由于训练数据规模较小导致的泛化能力较差的问题。在ATIS和Snips数据集上的实验结果表明,该模型能有效提升意图分类和槽填充的性能。 |
| 关 键 词： | 意图分类 槽填充 BERT 关联网络 |
BERT Based Joint Model for Intention Classification and Slot Filling |
|
| MA Tianyu,QIN Jun,LIU Jing,TIE Jun,HOU Qi.BERT Based Joint Model for Intention Classification and Slot Filling[J].Journal of Chinese Information Processing,2022,36(8):127-134. | |
| Authors: | MA Tianyu QIN Jun LIU Jing TIE Jun HOU Qi |
| Affiliation: | 1.School of Computer Science, South-Central University for Nationalities, Wuhan, Hubei 430074, China; 2.Hubei Provincial Engineering Research Center for Intelligent Management of Manufacturing Enterprises, Wuhan, Hubei 430074,China |
| Abstract: | The intention classification and the slot filling are two basic sub-tasks of spoken language understanding. A joint model of intention classification and slot filling based on BERT is proposed. Through an association network, the two tasks establish direct contact and share information. BERT is introduced into the model to enhance the semantic representation of word vectors, which effectively alleviates the issue of small training data. Experiments on ATIS and Snips data sets show that the proposed model can significantly improve the accuracy of intention classification and the F1 value of slot filling. |
| Keywords: | intent classification slot filling BERT association network |
| 点击此处可从《中文信息学报》浏览原始摘要信息 | |
| 点击此处可从《中文信息学报》下载全文 | |