Dialog

Best overviews: Chen et al 2017 - A Survey on Dialogue Systems: Recent Advances and New Frontiers and Roller et al 2020 - Recipes for building an open-domain chatbot.

• Introductions
  • S&LP - Ch 24
• General
  • Chen et al 2017 - A Survey on Dialogue Systems: Recent Advances and New Frontiers
  • Gao et al 2018 - Neural Approaches to Conversational AI (95 pages)
  • Roller et al 2020 - Open-Domain Conversational Agents: Current Progress, Open Problems, and Future Directions
  • Singh 2021 - A Survey on Near-human Conversational Agents
  • Caldarini et al 2022 - A Literature Survey of Recent Advances in Chatbots
  • Shaier et al 2022 - Mind the Knowledge Gap: A Survey of Knowledge-enhanced Dialogue Systems
  • Yi et al 2024 - A Survey on Recent Advances in LLM-Based Multi-turn Dialogue Systems
• Courses
  • Zhou Yu's Conversational AI course at Columbia
• Datasets
  • Serban et al 2017 - A Survey of Available Corpora for Building Data-Driven Dialogue Systems
  • Sundar & Heck 2022 - Multimodal Conversational AI A Survey of Datasets and Approaches
• Evaluation
  • Deriu et al 2019 - Survey on Evaluation Methods for Dialogue Systems
  • Fan et al 2020 - A Survey of Dialogue System Evaluation
  • Yeh et al 2021 - A Comprehensive Assessment of Dialog Evaluation Metrics
• Other
  • Ma et al 2020 - A Survey on Empathetic Dialogue Systems
  • Sundar & Heck 2022 - Multimodal Conversational AI: A Survey of Datasets and Approaches
  • Rahmani et al 2023 - A Survey on Asking Clarification Questions Datasets in Conversational Systems

• Ritter et al 2011 - Data-Driven Response Generation in Social Media
• Vinyals & Le 2015- A Neural Conversational Model
• Li et al 2016 - Deep Reinforcement Learning for Dialogue Generation
• Wolf et al 2019 - TransferTransfo: A Transfer Learning Approach for Neural Network Based Conversational Agents
• Li et al 2018 - Towards Deep Conversational Recommendations Introduced the ReDial dataset
• Meena: Adiwardana et al 2020 - Towards a Human-like Open-Domain Chatbot
• SOTA (2020, better than Meena) neural dialog generation: Roller et al 2020 - Recipes for building an open-domain chatbot. Code and models
• Andreas et al 2020 - Task-Oriented Dialogue as Dataflow Synthesis
• Dai et al 2022 - Dialog Inpainting: Turning Documents into Dialogs

See also Chatbots.

• Madotto et al 2020 - The Adapter-Bot: All-In-One Controllable Conversational Model

• Qian et al 2023 - Distinguish Sense from Nonsense: Out-of-Scope Detection for Virtual Assistants

• Prabhumoye 2020 - I love your chain mail! Making knights smile in a fantasy game world: Open-domain goal-oriented dialogue agents

• Bae et al 2022 - Building a Role Specified Open-Domain Dialogue System Leveraging Large-Scale Language Models

• Slonim et al 2021 - An Autonomous Debating System

See Vision and Language - Multimodal Dialog Agents.

For a nice overview, see related work section of Joshi 2024.

• Joshi et al 2024 - LLM-Based Open-Domain Integrated Task and Knowledge Assistants with Programmable Policies (Has a great overview)

A dialog act, also known as a speech act, is the “action” an utterance performs, such as request information, confirm, etc.

Dialog Acts are related to RST, except that dialog acts assume that the relation expressed in the dialog act is between adjacent utterances. See here for more information.

• Switchboard Dialog Act Draft 1997
• 2005 - Guidelines for Dialogue Act and Addressee Annotation Version 1.0 Guidelines for the AMI dataset
• Yu & Yu 2019 - MIDAS, A Dialog Act Annotation Scheme for Open-Domain Human-Machine Spoken Conversations There used to be a corpus and an open-source system available, but it has been taken down. Jurik uses this tagger in his system

• Paul et al 2019 - Towards Universal Dialogue Act Tagging for Task-Oriented Dialogues

Predicting the next dialog act. Papers:

• Tanaka et al 2019 - Dialogue-Act Prediction of Future Responses based on Conversation History

See also Learning from Feedback.

• Melamed & Niraula 2016 - Towards A Virtual Assistant That Can Be Taught New Tasks In Any Domain By Its End-Users Helpa paper
• Li et al 2016 - Learning through dialogue interactions by asking questions
• Labutov et al 2018 - LIA: A Natural Language Programmable Personal Assistant
• Li et al 2020 - Interactive Task Learning from GUI-Grounded Natural Language Instructions and Demonstrations
• Mininger 2021 - Expanding Task Diversity in Explanation-Based Interactive Task Learning

• Overviews
  • Keyvan et al 2023 - How to Approach Ambiguous Queries in Conversational Search: A Survey of Techniques, Approaches, Tools, and Challenges
    • Rahmani et al 2023 - A Survey on Asking Clarification Questions Datasets in Conversational Systems
• Papers
  • 2020 - Making Information Seeking Easier: An Improved Pipeline for Conversational Search
  • Dalton et al 2020 - CAsT-19: A Dataset for Conversational Information Seeking
• People
  • Jamie Callan

See also NLP progress - Dialog and Serban et al 2017 - A Survey of Available Corpora for Building Data-Driven Dialogue Systems.

• ReDial dataset - annotated dataset of dialogues, where users recommend movies to each other. Paper: Li et al 2018 - Towards Deep Conversational Recommendations
• MultiWOZ Also a data to text generation dataset. Paper v1.0: Ramadan et al 2018 - Large-Scale Multi-Domain Belief Tracking with Knowledge Sharing. Paper v2.0: Budzianowski et al 2018 - MultiWOZ - A Large-Scale Multi-Domain Wizard-of-Oz Dataset for Task-Oriented Dialogue Modelling
• Wizard of Wikipedia - Paper: Dinan et al 2018 - Wizard of Wikipedia: Knowledge-Powered Conversational Agents
• Self-dialogue Corpus - Paper: Fainberg et al 2018 - Talking to myself: self-dialogues as data for conversational agents
• Schema-Guided Dialogue Dataset - Paper: Rastogi et al 2019 - Towards Scalable Multi-domain Conversational Agents: The Schema-Guided Dialogue Dataset
• Topical-Chat - Paper: Gopalakrishnan et al 2019 - Topical-Chat: Towards Knowledge-Grounded Open-Domain Conversations
• Repository of Conversational Datasets - Paper: Henderson et al 2019 - A Repository of Conversational Datasets

• Apel et al 2023 - MeeQA: Natural Questions in Meeting Transcripts See the datasets they used for more multi-party datasets.

• Stolcke et al 2000 - Dialogue Act Modeling for Automatic Tagging and Recognition of Conversational Speech
• Boyer et al 2010 - Dialogue Act Modeling in a Complex Task-Oriented Domain
• Yu & Yu 2019 - MIDAS: A Dialog Act Annotation Scheme for Open Domain Human Machine Spoken Conversations

Safety: making sure the dialog agent doesn't use swear words, offensive language, give bad advice, etc, which could happen if the agent has been trained on a corpus of dialogs.

• Dinan et al 2019 - Build it Break it Fix it for Dialogue Safety: Robustness from Adversarial Human Attack
• Xu et al 2020 - Recipes for Safety in Open-domain Chatbots

• Perspective API

See also Chatbots - Evaluation.

• Walker et al 1997 - PARADISE: A Framework for Evaluating Spoken Dialogue Agents (Task oriented dialog I believe)

• Use dialog to add training data to improve the NER and EL on the fly. As examples get corrected, add them in. There should be prior work on this.

• Chatbots
  • Rule-based
    • FreeHAL wikipedia Rule-based, self-learning chatbot
  • Data-driven
• Task-Oriented Dialog
• Miscellaneous
  • LangChain Framework for building applications with prompting (chaining prompts, etc)

• SIGDIAL: 2020 2021
• IWSDS (International Workshop on Spoken Dialog Systems): 2023 2021 2017 Proceedings up to 2020 on dblp This is a nice workshop.

• Emily Dinan
• Maxine Eskenazi
• Andrea Madotto
• David Traum
• Marilyn Walker
• Jason Weston
• Zhou Yu

• Autonomous Language Agents
• Chatbots
• Discourse Analysis
• Embodied AI
• Multimodal Dialog Agents
• Task Oriented Dialog
• Jeff's List of Interesting Papers