EGOS-DST:对话现象感知和模式引导的一步对话状态追踪算法

doi:10.11896/jsjkx.210900246

摘要/Abstract

摘要： 为了平衡过度依赖本体和完全舍弃本体两种极端方式,近期的对话状态追踪工作专注于混合方式。目前,这些混合方式忽略了一些特殊现象,比如值共享和推荐接受。此外,被广泛使用的槽位门机制使模型很难并行处理槽位,并且还会将误差传播到槽值生成步骤。针对以上问题,提出一种新的混合方式,它能够处理多样性表达、未知值、值共享和推荐接受4种不同对话现象。通过修改候选值集合和模型输入,模型不再依赖槽位门机制并且能够一步并行处理槽位。实验结果显示,模型在英文数据集MultiWOZ 2.2和2.3上分别达到了57.7%和59.5%的联合目标准确率,在中文数据集RiSAWOZ上达到了68.1%,并且推理一次仅需10ms。最后还分析了模型的鲁棒性,在MultiWOZ 2.2上的结果显示即使推荐错误率达到15%,联合目标准确率仍有55.4%。

关键词: 任务导向对话系统, 对话状态追踪, BERT, 并行计算, 模式引导的对话

Abstract: Recent dialogue state tracking works have focused on the hybrid approach to balance the two extreme methods(i.e.,over-reliance on ontology and complete abandoning ontology).However,some special phenomena are ignored in these works.For instant,value sharing and recommendation acceptance.In addition,the widely used slot gate mechanism makes it difficult for the model to process slots in parallel and the mechanism also propagates errors to the slot value generation steps.This paper proposes a new hybrid approach that deals with four different phenomena,namely diverse value,unseen value,value sharing and recommendation acceptance.By modifying the candidate value set and the model input,our model can parallelly process slots in one step and no longer depend on the slot gate.Experimental results indicate that the model achieves 57.7% and 59.5% joint goal accuracy on the English dataset MultiWOZ 2.2 and 2.3,respectively,and reaches 68.1% on the Chinese dataset RiSAWOZ,with only 10ms infer time.Finally,the robustness of the model is analyzed.The results on MultiWOZ 2.2 show that the joint target accuracy rate is 55.4% when the recommendation error rate reaches 15%.

Key words: Task-oriented dialogue system, Dialogue state tracking, BERT, Parallel computing, Schema-guided dialogue

中图分类号:

TP391.1

朱若尘, 杨长春, 张登辉. EGOS-DST:对话现象感知和模式引导的一步对话状态追踪算法[J]. 计算机科学, 2022, 49(11A): 210900246-7. https://doi.org/10.11896/jsjkx.210900246

ZHU Ruo-chen, YANG Chang-chun, ZHANG Deng-hui. EGOS-DST:Efficient Schema-guided Approach to One-step Dialogue State Tracking for Diverse Expressions[J]. Computer Science, 2022, 49(11A): 210900246-7. https://doi.org/10.11896/jsjkx.210900246

参考文献

[1]RASTOGI A,ZANG X,SUNKARA S,et al.Schema-guided dialogue state tracking task at DSTC8[J].arXiv:2002.01359,2020.
[2]ZHONG V,XIONG C,SOCHER R.Global-Locally Self-Attentive Encoder for Dialogue State Tracking[C]//Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics(Volume 1:Long Papers).Melbourne:Association for Computational Linguistics,2018:1458-1467.
[3]SHAN Y,LI Z,ZHANG J,et al.A Contextual Hierarchical Attention Network with Adaptive Objective for Dialogue State Tracking[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.Online:Association for Computational Linguistics,2020:6322-6333.
[4]YE F,MANOTUMRUKSA J,ZHANG Q,et al.Slot Self-At-tentive Dialogue State Tracking[J].arXiv:2101.09374,2021.
[5]XU P,HU Q.An End-to-end Approach for Handling Unknown Slot Values in Dialogue State Tracking[C]//Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics(Volume 1:Long Papers).Melbourne:Association for Computational Linguistics,2018:1448-1457.
[6]CHAO G L,LANE I.Bert-dst:Scalable end-to-end dialoguestate tracking with bidirectional encoder representations from transformer[J].arXiv:1907.03040,2019.
[7]VINYALS O,FORTUNATO M,JAITLY N.Pointer Networks[C]//Advances in Neural Information Processing Systems.Montréal:Curran Associates,Inc.,2015:2692-2700.
[8]WANG Y,GUO Y,ZHU S.Slot Attention with Value Normalization for Multi-Domain Dialogue State Tracking[C]//Procee-dings of the 2020 Conference on Empirical Methods in Natural Language Processing(EMNLP).Online:Association for Computational Linguistics,2020:3019-3028.
[9]ZHU S,LI J,CHEN L,et al.Efficient Context and Schema Fusion Networks for Multi-Domain Dialogue State Tracking[C]//Findings of the Association for Computational Linguistics:EMNLP 2020.Online:Association for Computational Linguistics,2020:766-781.
[10]HAN T,LIU X,TAKANOBU R,et al.MultiWOZ 2.3:Amulti-domain task-oriented dataset enhanced with annotation corrections and co-reference annotation[J].arXiv:2010.05594,2020.
[11]ZANG X,RASTOGI A,SUNKARA S,et al.MultiWOZ 2.2:A Dialogue Dataset with Additional Annotation Corrections and State Tracking Baselines[C]//Proceedings of the 2nd Workshop on Natural Language Processing for Conversational AI.Online:Association for Computational Linguistics,2020:109-117.
[12]OUYANG Y,CHEN M,DAI X,et al.Dialogue State Tracking with Explicit Slot Connection Modeling[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.Online:Association for Computational Linguistics,2020:34-40.
[13]HECK M,NIEKERK C,LUBIS N,et al.TripPy:A Triple Copy Strategy for Value Independent Neural Dialog State Tracking[C]//Proceedings of the 21th Annual Meeting of the Special Interest Group on Discourse and Dialogue.1st virtual meeting:Association for Computational Linguistics,2020:35-44.
[14]HENDERSON M,THOMSON B,YOUNG S.Word-Based Dialog State Tracking with Recurrent Neural Networks[C]//Proceedings of the 15th Annual Meeting of the Special Interest Group on Discourse and Dialogue(SIGDIAL).Philadelphia,PA:Association for Computational Linguistics,2014:292-299.
[15]ZILKA L,JURCICEK F.Incremental LSTM-based dialog state tracker[C]//2015 IEEE Workshop on Automatic Speech Recognition and Understanding(ASRU).Scottsdale,AZ:IEEE Press,2015:757-762.
[16]MRKŠIĆ N,Ó SÉAGHDHA D,WEN T H,et al.Neural Belief Tracker:Data-Driven Dialogue State Tracking[C]//Proceedings of the 55th Annual Meeting of the Association for Computa-tional Linguistics(Volume 1:Long Papers).Vancouver:Association for Computational Linguistics,2017:1777-1788.
[17]WU C-S,MADOTTO A,HOSSEINI-ASL E,et al.Transferable Multi-Domain State Generator for Task-Oriented Dialogue Systems[C]//Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics.Florence:Association for Computational Linguistics,2019:808-819.
[18]GOEL R,PAUL S,HAKKANI-TÜR D.Hyst:A hybrid ap-proach for flexible and accurate dialogue state tracking[J].ar-Xiv:1907.00883,2019.
[19]RASTOGI A,ZANG X,SUNKARA S,et al.Towards scalable multi-domain conversational agents:The schema-guided dialogue dataset[C]//Proceedings of the AAAI Conference on Artificial Intelligence.New York:AAAI Press,2020:8689-8696.
[20]ZHANG J,HASHIMOTO K,WU C S,et al.Find or Classify? Dual Strategy for Slot-Value Predictions on Multi-Domain Dialog State Tracking[C]//Proceedings of the Ninth Joint Conference on Lexical and Computational Semantics.Barcelona,Spain(Online):Association for Computational Linguistics,2020:154-167.
[21]KIM S,YANG S,KIM G,et al.Efficient Dialogue State Tracking by Selectively Overwriting Memory[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.Online:Association for Computational Linguistics,2020:567-582.
[22]ZENG Y,NIE J Y.Multi-domain dialogue state tracking based on state graph[J].arXiv:2010.11137,2020.
[23]HU J,YANG Y,CHEN C,et al.SAS:Dialogue State Tracking via Slot Attention and Slot Information Sharing[C]//Procee-dings of the 58th Annual Meeting of the Association for Computational Linguistics.Online:Association for Computational Linguistics,2020:6366-6375.
[24]VASWANI A,SHAZEER N,PARMAR N,et al.Attention isAll you Need[C]//Advances in Neural Information Processing Systems.Long Beach,CA:Curran Associates,Inc.,2017:5998-6008.
[25]DEVLIN J,CHANG M W,LEE K,et al.BERT:Pre-training of Deep Bidirectional Transformers for Language Understanding[C]//Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics:Human Language Technologies,Volume 1(Long and Short Papers).Minneapolis,Minnesota:Association for Computational Linguistics,2019:4171-4186.
[26]WU C S,HOI S C H,SOCHER R,et al.TOD-BERT:Pre-trained Natural Language Understanding for Task-Oriented Dialogue[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing(EMNLP).Online:Association for Computational Linguistics,2020:917-929.
[27]QUAN J,ZHANG S,CAO Q,et al.RiSAWOZ:A Large-Scale Multi-Domain Wizard-of-Oz Dataset with Rich Semantic Annotations for Task-Oriented Dialogue Modeling[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing(EMNLP).Online:Association for Computational Linguistics,2020:930-940.
[28]BOWMAN S R,VILNIS L,VINYALS O,et al.Generating Sentences from a Continuous Space[C]//Proceedings of The 20th SIGNLL Conference on Computational Natural Language Learning.Berlin:Association for Computational Linguistics,2016:10-21.
[29]KINGMA D P,BA J.Adam:A method for stochastic optimization[J].arXiv:1412.6980,2014.
[30]HOWARD J,RUDER S.Universal Language Model Fine-tuning for Text Classification[C]//Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics(Volume 1:Long Papers).Melbourne:Association for Computational Linguistics,2018:328-339.
[31]REN L,NI J,MCAULEY J.Scalable and Accurate DialogueState Tracking via Hierarchical Sequence Generation[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing(EMNLP-IJCNLP).Hong Kong:Association for Computational Linguistics,2019:1876-1885.
[32]ZHOU L,SMALL K.Multi-domain dialogue state tracking as dynamic knowledge graph enhanced question answering[J].arXiv:1911.06192,2019.
[33]LEE H,LEE J,KIM T Y.SUMBT:Slot-Utterance Matching for Universal and Scalable Belief Tracking[C]//Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics.Florence:Association for Computational Linguistics,2019:5478-5483.
[34]FENG Y,WANG Y,LI H.A Sequence-to-Sequence Approach to Dialogue State Tracking[J].arXiv:2011.09553,2020.
[35]QUAN J,XIONG D.Modeling Long Context for Task-Oriented Dialogue State Generation[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.Online:Association for Computational Linguistics,2020:7119-7124.
[36]SANH V,DEBUT L,CHAUMOND J,et al.DistilBERT,a distilled version of BERT:smaller,faster,cheaper and lighter[J].arXiv:1910.01108,2019.

相关文章 15

[1]	于家畦, 康晓东, 白程程, 刘汉卿. 一种新的中文电子病历文本检索模型 New Text Retrieval Model of Chinese Electronic Medical Records 计算机科学, 2022, 49(6A): 32-38. https://doi.org/10.11896/jsjkx.210400198
[2]	康雁, 吴志伟, 寇勇奇, 张兰, 谢思宇, 李浩. 融合Bert和图卷积的深度集成学习软件需求分类 Deep Integrated Learning Software Requirement Classification Fusing Bert and Graph Convolution 计算机科学, 2022, 49(6A): 150-158. https://doi.org/10.11896/jsjkx.210500065
[3]	余本功, 张子薇, 王惠灵. 一种融合多层次情感和主题信息的TS-AC-EWM在线商品排序方法 TS-AC-EWM Online Product Ranking Method Based on Multi-level Emotion and Topic Information 计算机科学, 2022, 49(6A): 165-171. https://doi.org/10.11896/jsjkx.210400238
[4]	陈鑫, 李芳, 丁海昕, 孙唯哲, 刘鑫, 陈德训, 叶跃进, 何香. 面向国产异构众核架构的CFD非结构网格计算并行优化方法 Parallel Optimization Method of Unstructured-grid Computing in CFD for DomesticHeterogeneous Many-core Architecture 计算机科学, 2022, 49(6): 99-107. https://doi.org/10.11896/jsjkx.210400157
[5]	郭雨欣, 陈秀宏. 融合BERT词嵌入表示和主题信息增强的自动摘要模型 Automatic Summarization Model Combining BERT Word Embedding Representation and Topic Information Enhancement 计算机科学, 2022, 49(6): 313-318. https://doi.org/10.11896/jsjkx.210400101
[6]	韦入铭, 陈若愚, 李晗, 刘旭红. 基于深度学习与文本计量的技术趋势分析 Analysis of Technology Trends Based on Deep Learning and Text Measurement 计算机科学, 2022, 49(11A): 211100119-6. https://doi.org/10.11896/jsjkx.211100119
[7]	陈孜卓, 林夕, 王中卿. 基于论据边界识别的立场分类研究 Stance Detection Based on Argument Boundary Recognition 计算机科学, 2022, 49(11A): 210800180-5. https://doi.org/10.11896/jsjkx.210800180
[8]	黄佳为, 李晓鹏, 凌诚. 一种基于GPU的核苷酸分子系统发育树条件似然概率可扩展并行计算方法 Scalable Parallel Computing Method for Conditional Likelihood Probability of Nucleotide Molecular Phylogenetic Tree Based on GPU 计算机科学, 2022, 49(11A): 210800189-7. https://doi.org/10.11896/jsjkx.210800189
[9]	程思伟, 葛唯益, 王羽, 徐建. BGCN:基于BERT和图卷积网络的触发词检测 BGCN:Trigger Detection Based on BERT and Graph Convolution Network 计算机科学, 2021, 48(7): 292-298. https://doi.org/10.11896/jsjkx.200500133
[10]	傅天豪, 田鸿运, 金煜阳, 杨章, 翟季冬, 武林平, 徐小文. 一种面向构件化并行应用程序的性能骨架分析方法 Performance Skeleton Analysis Method Towards Component-based Parallel Applications 计算机科学, 2021, 48(6): 1-9. https://doi.org/10.11896/jsjkx.201200115
[11]	何亚茹, 庞建民, 徐金龙, 朱雨, 陶小涵. 基于神威平台的Floyd并行算法的实现和优化 Implementation and Optimization of Floyd Parallel Algorithm Based on Sunway Platform 计算机科学, 2021, 48(6): 34-40. https://doi.org/10.11896/jsjkx.201100051
[12]	董哲, 邵若琦, 陈玉梁, 翟维枫. 基于BERT和对抗训练的食品领域命名实体识别 Named Entity Recognition in Food Field Based on BERT and Adversarial Training 计算机科学, 2021, 48(5): 247-253. https://doi.org/10.11896/jsjkx.200800181
[13]	冯凯, 马鑫玉. (n,k)-冒泡排序网络的子网络可靠性 Subnetwork Reliability of (n,k)-bubble-sort Networks 计算机科学, 2021, 48(4): 43-48. https://doi.org/10.11896/jsjkx.201100139
[14]	胡蓉, 阳王东, 王昊天, 罗辉章, 李肯立. 基于GPU加速的并行WMD算法 Parallel WMD Algorithm Based on GPU Acceleration 计算机科学, 2021, 48(12): 24-28. https://doi.org/10.11896/jsjkx.210600213
[15]	陈德, 宋华珠, 张娟, 周泓林. 融合BERT和记忆网络的实体识别 Entity Recognition Fusing BERT and Memory Networks 计算机科学, 2021, 48(10): 91-97. https://doi.org/10.11896/jsjkx.200900015

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed