云社区 博客 博客详情
云社区 博客 博客详情

ACL 2020对话系统相关论文整理

listen2Bot 发表于 2020-05-29 09:32:40 05-29 09:32
listen2Bot 发表于 2020-05-29 09:32:40 2020/05/29
0
1

【摘要】 ACL 2020 总共收录56篇Dialogue相关的论文,达到历史之最。其中对话生成相关的论文数量最多达到15篇,其次是对话新任务和数据的发布大概有7篇,对话系统评价分析相关7篇,对话状态跟踪7篇,端到端对话系统6篇,对话理解相关5篇,对话策略学习4篇,多模态相关4篇,其他1篇。

ACL 2020 总共收录56Dialogue相关的论文,达到历史之最。其中对话生成相关的论文数量最多达到15篇,其次是对话新任务和数据的发布大概有7篇,对话系统评价分析相关7篇,对话状态跟踪7篇,端到端对话系统6篇,对话理解相关5篇,对话策略学习4篇,多模态相关4篇,其他1篇。后续将会对其中重点论文进行分享总结。

对话理解

1.   Recursive Template-based Frame Generation for Task Oriented Dialog

2.   Span-ConveRT: Few-shot Span Extraction for Dialog with Pretrained Conversational Representations

3.   Coach: A Coarse-to-Fine Approach for Cross-domain Slot Filling

4.   Few-shot Slot Tagging with Collapsed Dependency Transfer and Label-enhanced Task-adaptive Projection Network

5.   Unknown Intent Detection Using Gaussian Mixture Model with an Application to Zero-shot Intent Classification

对话状态跟踪

1.   A Contextual Hierarchical Attention Network with Adaptive Objective for Dialogue State Tracking

2.   Efficient Dialogue State Tracking by Selectively Overwriting Memory

3.   Rethinking Dialogue State Tracking with Reasoning

4.   SAS: Dialogue State Tracking via Slot Attention and Slot Information Sharing

5.   Zero-Shot Transfer Learning with Synthesized Data for Multi-Domain Dialogue State Tracking

6.   Dialogue State Tracking with Explicit Slot Connection Modeling

7.   Modeling Long Context for Task-Oriented Dialogue State Generation

对话策略

1.   Learning Dialog Policies from Weak Demonstrations

2.   Learning Efficient Dialogue Policy from Demonstrations through Shaping

3.   Multi-Agent Task-Oriented Dialog Policy Learning with Role-Aware Reward Decomposition

4.   Semi-Supervised Dialogue Policy Learning via Stochastic Reward Estimation

对话生成

1.   Data Manipulation: Towards Effective Instance Learning for Neural Dialogue Generation via Learning to Augment and Reweight

2.   Diverse and Informative Dialogue Generation with Context-Specific Commonsense Knowledge Awareness

3.   Diversifying Dialogue Generation with Non-Conversational Text

4.   Generate, Delete and Rewrite: A Three-Stage Framework for Improving Persona Consistency of Dialogue Generation

5.   Learning to Customize Model Structures for Few-shot Dialogue Generation Tasks

6.   Multi-Domain Dialogue Acts and Response Co-Generation

7.   Negative Training for Neural Dialogue Response Generation

8.   Neural Generation of Dialogue Response Timings

9.   Paraphrase Augmented Task-Oriented Dialog Generation

10.   PLATO: Pre-trained Dialogue Generation Model with Discrete Latent Variable

11.   Slot-consistent NLG for Task-oriented Dialogue Systems with Iterative Rectification Network

12.   Towards Holistic and Automatic Evaluation of Open-Domain Dialogue Generation

13.    USR: An Unsupervised and Reference Free Evaluation Metric for Dialog Generation

14.    You Impress Me: Dialogue Generation via Mutual Persona Perception

15.  None of the Above”: Measure Uncertainty in Dialog Response Retrieval

端到端对话系统

1.   Dynamic Fusion Network for Multi-Domain End-to-end Task-Oriented Dialog

2.   End-to-End Neural Pipeline for Goal-Oriented Dialogue Systems using GPT-2

3.   Grounding Conversations with Improvised Dialogues

4.   Large Scale Multi-Actor Generative Dialog Modeling

5.   Conversational Word Embedding for Retrieval-Based Dialog System

6.   Learning Low-Resource End-To-End Goal-Oriented Dialog for Fast and Reliable System Deployment

对话系统评价分析

1.   Beyond User Self-Reported Likert Scale Ratings: A Comparison Model for Automatic Dialog Evaluation

2.   Dialogue Coherence Assessment Without Explicit Dialogue Act Labels

3.   Don’t Say That! Making Inconsistent Dialogue Unlikely with Unlikelihood Training

4.   Crawling and Preprocessing Mailing Lists At Scale for Dialog Analysis

5.   Designing Precise and Robust Dialogue Response Evaluators

6.   Evaluating Dialogue Generation Systems via Response Selection

7.   Learning an Unreferenced Metric for Online Dialogue Evaluation

多模态对话

1.   History for Visual Dialog: Do we really need it?

2.   The Dialogue Dodecathlon: Open-Domain Knowledge and Image Grounded Conversational Agents

3.   Towards Emotion-aided Multi-modal Dialogue Act Classification

4.   Video-Grounded Dialogues with Pretrained Generation Language Models

对话新任务/数据

1.   KdConv: A Chinese Multi-domain Dialogue Dataset Towards Multi-turn Knowledge-driven Conversation

2.   More Diverse Dialogue Datasets via Diversity-Informed Data Collection

3.   MuTual: A Dataset for Multi-Turn Dialogue Reasoning

4.   Storytelling with Dialogue: A Critical Role Dungeons and Dragons Dataset

5.   ChartDialogs: Plotting from Natural Language Instructions

6.   MIE: A Medical Information Extractor towards Medical Dialogues

7.   Towards Conversational Recommendation over Multi-Type Dialogs

其他

1.   Learning to execute instructions in a Minecraft dialogue


登录后可下载附件,请登录或者注册

【版权声明】本文为华为云社区用户原创内容,转载时必须标注文章的来源(华为云社区),文章链接,文章作者等基本信息,否则作者和本社区有权追究责任。如果您发现本社区中有涉嫌抄袭的内容,欢迎发送邮件至:huaweicloud.bbs@huawei.com进行举报,并提供相关证据,一经查实,本社区将立刻删除涉嫌侵权内容。
评论文章 //点赞 收藏 1
点赞
分享文章到微博
分享文章到朋友圈

评论 (0)


0/1000
评论

登录后可评论,请 登录注册

评论

您还没有写博客的权限!

温馨提示

您确认删除评论吗?

确定
取消
温馨提示

您确认删除评论吗?

删除操作无法恢复,请谨慎操作。

确定
取消
温馨提示

您确认删除博客吗?

确定
取消

确认删除

您确认删除博客吗?

确认删除

您确认删除评论吗?

温馨提示

登录超时或用户已下线,请重新登录!!!

确定
取消