Skip to content

A Paper List for Open-Domain Dialogue Generation, and related datasets.

Notifications You must be signed in to change notification settings

csnlp/Dialogue-Generation

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

23 Commits
 
 

Repository files navigation

This repository is to document the papers focusing on dialogue generation. I will also include their code implementation if possible.

Contents

Open-Domain Dialogue Generation

Open-Domain Dialogue Datasets

  1. The Ubuntu Dialogue Corpus, SIGDIAL: paper, dataset
  2. Cornell Movie Dialogs Corpus, dataset
  3. A Survey Paper about Datasets of Collection from McGill & UdeM, dataset
  4. NLPCC 2017(chinese): Emotional Dialogue Generation, datasets
  5. OpenSubtitles2016: Extracting Large Parallel Corpora from Movie and TV Subtitles, LREC 2016. paper, dataset
  6. DailyDialog, IJCNLP. paper, dataset
  7. Twitter Customer Service. dataset
  8. Douban Corpus(chinese), ACL 2017. dataset, paper

Seq2Seq Model for Dialogue Generation

Actually, almost all of the existing generative approach use the Seq2Seq model.

  1. Neural Responding Machine for Short-Text Conversation, ACL 2015. paper.
  2. A Neural Conversational Model, arxiv. paper

Personalized Dialogue Generation

  1. A Pre-training Based Personalized Dialogue Generation Model with Persona-sparse Data, AAAI 2020. paper
  2. Guiding Variational Response Generator to Exploit Persona, ACL 2020. paper

Reinforcement Learning Approaches for Dialogue Generation

  1. Deep Reinforcement Learning for Dialogue Generation, paper
  2. Hierarchical Reinforcement Learning for Open-Domain Dialog, AAAI 2020. paper

Adversarial Learning Approaches for Dialogue Generation

  1. Adversarial Learning for Neural Dialogue Generation, paper

Context-aware Dialogue Generation:Hierarchy Methods

  1. A Hierarchical Recurrent Encoder-Decoder for Generative Context-Aware Query Suggestion, the first HRED model, CIKM 2015. paper
  2. Building End-To-End Dialogue Systems Using Generative Hierarchical Neural Network Models, the famous HRED model. paper.
  3. A Hierarchical Latent Variable Encoder-Decoder Model for Generating Dialogues, VHRED model, AAAI 2017. paper.
  4. Hierarchical Variational Memory Network for Dialogue Generation,WWW 2018. paper

Evaluation Methods

  1. How NOT To Evaluate Your Dialogue System: An Empirical Study of Unsupervised Evaluation Metrics for Dialogue Response Generation, A empirical study. paper
  2. Predictive Engagement: An Efficient Metric for Automatic Evaluation of Open-Domain Dialogue Systems, AAAI 2020. paper.
  3. Learning from Easy to Complex: Adaptive Multi-curricula Learning for Neural Dialogue Generation, AAAI 2020.
  4. Towards Holistic and Automatic Evaluation of Open-Domain Dialogue Generation, ACL 2020. paper
  5. USR: An Unsupervised and Reference Free Evaluation Metric for Dialog Generation, ACL 2020. paper
  6. Evaluating Dialogue Generation Systems via Response Selection, ACL 2020. paper

Pretrained Model for Dialogue Generation

  1. DIALOGPT : Large-Scale Generative Pre-training for Conversational Response Generation, ACL 2020. paper.

Papers about The Safe Response Problem

  1. A Diversity-Promoting Objective Function for Neural Conversation Models(MMI), NAACL-HLT 2016. paper
  2. Learning Discourse-level Diversity for Neural Dialog Models using Conditional Variational Autoencoders, ACL 2017. paper
  3. Topic Aware Neural Response Generation, AAAI 2017. paper
  4. Generating Informative and Diverse Conversational Responses via Adversarial Information Maximization, NIPS 2018. paper
  5. Diversifying Dialogue Generation with Non-Conversational Text, ACL 2020. paper.
  6. DAL: Dual Adversarial Learning for Dialogue Generation, NAACL 2019. paper
  7. Negative Training for Neural Dialogue Response Generation, ACL 2020. paper
  8. Generating Informative Conversational Response using Recurrent Knowledge-Interaction and Knowledge-Copy

Incoporate Knowledge into Response Generation

  1. Conversational Graph Grounded Policy Learning for Open-Domain Conversation Generation, ACL 2020. paper
  2. Diverse and Informative Dialogue Generation with Context-Specific Commonsense Knowledge Awareness, ACL 2020. paper
  3. Grounded Conversation Generation as Guided Traverses in Commonsense Knowledge Graphs, ACL 2020. paper
  4. Response-Anticipated Memory for On-Demand Knowledge Integration in Response Generation, ACL 2020. paper

Improve Dialogue Consistency

  1. Generate, Delete and Rewrite: A Three-Stage Framework for Improving Persona Consistency of Dialogue Generation, ACL 2020. paper

Curriculum Learning Approaches

  1. Learning from Easy to Complex: Adaptive Multi-curricula Learning for Neural Dialogue Generation, AAAI 2020.
  2. CDL: Curriculum Dual Learning for Emotion-Controllable Response Generation, ACL 2020.

Incorporate External Knowledge into Dialogue Generation

  1. Improving Knowledge-aware Dialogue Generation via Knowledge Base Question Answering, AAAI 2020. paper

Other Types

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

Task-oriented Dialogue Generation

Datasets for Task-oriented Dialogue Systems

  1. FRAMES Corpus. paper, dataset
  2. Towards Scalable Multi-Domain Conversational Agents: The Schema-Guided Dialogue Dataset, AAAI 2020. paper. A large dataset containing over 16k multi-domain conversations spanning 16 domains.

Multi-Domain Dialogue Generation

  1. MALA: Cross-Domain Dialogue Generation with Action Learning, AAAI 2020. paper
  2. Multi-Domain Dialogue Acts and Response Co-Generation,ACL 2020. paper

End-to-End Approaches

  1. MOSS: End-to-End Dialog System Framework with Modular Supervision, paper

Others

  1. End-to-End Trainable Non-Collaborative Dialog System, AAAI 2020. paper

Academic Conferences and Workshops

  1. ACL 20xx
  2. EMNLP 20xx
  3. NAACL-HLT 20xx
  4. COLING 20xx
  5. SIGDIAL 20xx

Releases

No releases published

Packages

No packages published