The Design and Implementation of XiaoIce, an Empathetic Social Chatbot
Not on the map. Why?
Abstract (arXiv)
This paper describes the development of Microsoft XiaoIce, the most popular social chatbot in the world. XiaoIce is uniquely designed as an AI companion with an emotional connection to satisfy the human need for communication, affection, and social belonging. We take into account both intelligent quotient (IQ) and emotional quotient (EQ) in system design, cast human-machine social chat as decision-making over Markov Decision Processes (MDPs), and optimize XiaoIce for long-term user engagement, measured in expected Conversation-turns Per Session (CPS). We detail the system architecture and key components including dialogue manager, core chat, skills, and an empathetic computing module. We show how XiaoIce dynamically recognizes human feelings and states, understands user intent, and responds to user needs throughout long conversations. Since her launch in 2014, XiaoIce has communicated with over 660 million active users and succeeded in establishing long-term relationships with many of them. Analysis of large scale online logs shows that XiaoIce has achieved an average CPS of 23, which is significantly higher than that of other chatbots and even human conversations.
Key facts. Cited by 172 papers in the map (data through 7 Oct 2026). First posted on arXiv 21 Dec 2018.
Builds on
No reference list yet.
Built on by
Citing papers per year: 2019: 1; 2020: 2; 2021: 36; 2022: 29; 2023: 37; 2024: 30; 2025: 23; 2026: 14.
172 papers.
Sort:
By area:
- 2024Evaluating Very Long-Term Conversational Memory of LLM Agentscited by 800
- 2021Towards Emotional Support Dialog Systemscited by 294
- 2019Way Off-Policy Batch Deep Reinforcement Learning of Implicit Human Preferences in Dialogcited by 283
- 2023A Comprehensive Survey of AI-Generated Content (AIGC): A History of Generative AI from GAN to ChatGPTcited by 221
- 2022BlenderBot 3: a deployed conversational agent that continually learns to responsibly engagecited by 197
- 2022Pretrained Language Models for Text Generation: A Surveycited by 121
- 2022ProsocialDialog: A Prosocial Backbone for Conversational Agentscited by 109
- 2020Human-centric Dialog Training via Offline Reinforcement Learningcited by 92
- 2023A Complete Survey on Generative AI (AIGC): Is ChatGPT from GPT-4 to GPT-5 All You Need?cited by 84
- 2021DialogueCRN: Contextual Reasoning Networks for Emotion Recognition in Conversationscited by 71
- 2024A Complete Survey on LLM-based AI Chatbotscited by 71
- 2020Dialogue Response Ranking Training with Large-Scale Human Feedback Datacited by 68
38 more
- 2022Control Globally, Understand Locally: A Global-to-Local Hierarchical Graph Network for Emotional Support Conversationcited by 56
- 2023Is ChatGPT Equipped with Emotional Dialogue Capabilities?cited by 54
- 2023Leveraging Large Language Models to Power Chatbots for Collecting User Self-Reported Datacited by 48
- 2021Biomedical Question Answering: A Survey of Approaches and Challengescited by 46
- 2023Faithful Persona-based Conversational Dataset Generation with Large Language Modelscited by 45
- 2021CoMAE: A Multi-factor Hierarchical Framework for Empathetic Response Generationcited by 42
- 2024A Full-duplex Speech Dialogue Scheme Based On Large Language Modelscited by 40
- 2021One Chatbot Per Person: Creating Personalized Chatbots based on Implicit User Profilescited by 39
- 2021PLATO-XL: Exploring the Large-scale Pre-training of Dialogue Generationcited by 37
- 2023Are Personalized Stochastic Parrots More Dangerous? Evaluating Persona Biases in Dialogue Systemscited by 37
- 2023A Transformer-Based Model With Self-Distillation for Multimodal Emotion Recognition in Conversationscited by 37
- 2024Large Language Models Produce Responses Perceived to be Empathiccited by 36
- 2021ValueNet: A New Dataset for Human Value Driven Dialogue Systemcited by 34
- 2021Revealing Persona Biases in Dialogue Systemscited by 33
- 2021Knowledge-Grounded Dialogue Generation with a Unified Knowledge Representationcited by 32
- 2023Foundation Metrics for Evaluating Effectiveness of Healthcare Conversations Powered by Generative AIcited by 32
- 2023Artificial Influence: An Analysis Of AI-Driven Persuasioncited by 31
- 2024Measuring and Benchmarking Large Language Models' Capabilities to Generate Persuasive Languagecited by 31
- 2022Duplex Conversation: Towards Human-like Interaction in Spoken Dialogue Systemscited by 28
- 2023LiveChat: A Large-Scale Personalized Dialogue Dataset Automatically Constructed from Live Streamingcited by 27
- 2022A Literature Survey of Recent Advances in Chatbotscited by 26
- 2024Building Better AI Agents: A Provocation on the Utilisation of Persona in LLM-based Conversational Agentscited by 26
- 2025Enhancing Personalized Multi-Turn Dialogue with Curiosity Rewardcited by 25
- 2023Using Large Language Models for Zero-Shot Natural Language Generation from Knowledge Graphscited by 24
- 2024Jill Watson: A Virtual Teaching Assistant powered by ChatGPTcited by 24
- 2023Helping the Helper: Supporting Peer Counselors via AI-Empowered Practice and Feedbackcited by 23
- 2023Bias and Fairness in Chatbots: An Overviewcited by 20
- 2024CloChat: Understanding How People Customize, Interact, and Experience Personas in Large Language Modelscited by 19
- 2022FADO: Feedback-Aware Double COntrolling Network for Emotional Support Conversationcited by 18
- 2023Emotion Rendering for Conversational Speech Synthesis with Heterogeneous Graph-Based Context Modelingcited by 17
- 2021OpenViDial 2.0: A Larger-Scale, Open-Domain Dialogue Generation Dataset with Visual Contextscited by 16
- 2022HeterMPC: A Heterogeneous Graph Neural Network for Response Generation in Multi-Party Conversationscited by 16
- 2022Multimodal Conversational AI: A Survey of Datasets and Approachescited by 16
- 2022A Mixture-of-Expert Approach to RL-based Dialogue Managementcited by 16
- 2023Let's have a chat! A Conversation with ChatGPT: Technology, Applications, and Limitationscited by 16
- 2021Partner Matters! An Empirical Study on Fusing Personas for Personalized Response Selection in Retrieval-Based Chatbotscited by 15
- 2021Automatically Select Emotion for Response via Personality-affected Emotion Transitioncited by 15
- 2022Empathetic Conversational Systems: A Review of Current Advances, Gaps, and Opportunitiescited by 14
50 most cited of 172 All 172 (JSON)
- 2026Neither Silence nor Overlap Is Failure: Intent-Conditioned Evaluation of Turn-Taking in Full-Duplex Spoken Dialogue Models23 Sep 2026
- 2026DeepSAGE: Stage-Aware Reinforcement Learning for Structured CBT Counseling Dialogue23 Aug 2026
- 2026S2Dialog: Multimodal Dialogue Retrieval with Semantic and Acoustic-Style Modeling14 Aug 2026
- 2026I wanted it to feel more personal: Customization of social AI as AI individualism in practice20 Jul 2026
- 2026AuEmoChat: Authentic Emotion Understanding and Rendering for Conversational Speech Synthesis17 Jul 2026
- 2026Adaptive Turn-Taking for Real-time Multi-Party Voice Agents11 Jun 2026
- 2026SURE: Synergistic Uncertainty-aware Reasoning for Multimodal Emotion Recognition in Conversations2 Apr 2026
- 2026Dynamic Fusion-Aware Graph Convolutional Neural Network for Multimodal Emotion Recognition in Conversations22 Mar 2026
- 2026CharacterFlywheel: Scaling Iterative Improvement of Engaging and Steerable LLMs in Production2 Mar 2026
- 2026Hear You in Silence: Designing for Active Listening in Human Interaction with Conversational Agents Using Context-Aware Pacing5 Feb 2026
- 2026Privacy in Human-AI Romantic Relationships: Concerns, Boundaries, and Agency23 Jan 2026
- 2026Scaling Ambiguity: Augmenting Human Annotation in Speech Emotion Recognition with Audio-Language Models21 Jan 2026
38 more
- 2026MASCOT: Multi-Agent Socio-Collaborative Companion Systems20 Jan 2026
- 2026Do You Understand How I Feel?: Towards Verified Empathy in Therapy Chatbots13 Jan 2026
- 2025Future You: Designing and Evaluating Multimodal AI-generated Digital Twins for Strengthening Future Self-Continuity5 Dec 2025
- 2025Investigating AI in Peer Support via Multi-Module System-Driven Embodied Conversational Agents27 Nov 2025
- 2025Harmful Traits of AI Companions18 Nov 2025
- 2025Systematizing LLM Persona Design: A Four-Quadrant Technical Taxonomy for AI Companion Applications4 Nov 2025
- 2025Mutual Wanting in Human--AI Interaction: Empirical Evidence from Large-Scale Analysis of GPT Model Transitions27 Oct 2025
- 2025Higher Satisfaction, Lower Cost: A Technical Report on How LLMs Revolutionize Meituan's Intelligent Interaction Systems15 Oct 2025
- 2025A Single Character can Make or Break Your LLM Evals2 Oct 2025
- 2025Mental Health Impacts of AI Companions: Triangulating Social Media Quasi-Experiments, User Perspectives, and Relational Theory26 Sep 2025
- 2025E3RG: Building Explicit Emotion-driven Empathetic Response Generation System with Multimodal Large Language Model18 Aug 2025
- 2025In-Context Examples Matter: Improving Emotion Recognition in Conversation with Instruction Tuning16 Aug 2025
- 2025Understanding User Preferences for Interaction Styles in Conversational Recommender Systems: The Predictive Role of System Qualities, User Experience, and Traits4 Aug 2025
- 2025Evaluating User Experience in Conversational Recommender Systems: A Systematic Review Across Classical and LLM-Powered Approaches4 Aug 2025
- 2025Toward Real-World Chinese Psychological Support Dialogues: CPsDD Dataset and a Co-Evolving Multi-Agent System10 Jul 2025
- 2025WildSpeech-Bench: Benchmarking End-to-End SpeechLLMs in the Wild27 Jun 2025
- 2025MEDAL: A Framework for Benchmarking LLMs as Multilingual Open-Domain Dialogue Evaluators28 May 2025
- 2025PaRT: Enhancing Proactive Social Chatbots with Personalized Real-Time Retrieval29 Apr 2025
- 2025Balancing Task-invariant Interaction and Task-specific Adaptation for Unified Image Fusion7 Apr 2025
- 2025Enhancing Personalized Multi-Turn Dialogue with Curiosity Reward4 Apr 2025
- 2025KwaiChat: A Large-Scale Video-Driven Multilingual Mixed-Type Dialogue Corpus10 Mar 2025
- 2025LLM-Enhanced Dialogue Management for Full-Duplex Spoken Dialogue Systems19 Feb 2025
- 2025A Taxonomy of Linguistic Expressions That Contribute To Anthropomorphism of Language Technologies14 Feb 2025
- 2025Mind the Gap! Choice Independence in Using Multilingual LLMs for Persuasive Co-Writing Tasks in Different Languages13 Feb 2025
- 2025Retrieval-Augmented Dialogue Knowledge Aggregation for Expressive Conversational Speech Synthesis11 Jan 2025
- 2024Intra- and Inter-modal Context Interaction Modeling for Conversational Speech Synthesis25 Dec 2024
- 2024CMATH: Cross-Modality Augmented Transformer with Hierarchical Variational Distillation for Multimodal Emotion Recognition in Conversation15 Nov 2024
- 2024Wild Narratives: Exploring the Effects of Animal Chatbots on Empathy and Positive Attitudes toward Animals9 Nov 2024
- 2024Understanding Communication Preferences of Information Workers in Engagement with Text-Based Conversational Agents27 Oct 2024
- 2024Decoding Hate: Exploring Language Models' Reactions to Hate Speech1 Oct 2024
- 2024Self-Emotion Blended Dialogue Generation in Social Simulation Agents3 Aug 2024
- 2024GermanPartiesQA: Benchmarking Commercial Large Language Models and AI Companions for Political Alignment and Sycophancy25 Jul 2024
- 2024Exploring Human-LLM Conversations: Mental Models and the Originator of Toxicity8 Jul 2024
- 2024Measuring and Benchmarking Large Language Models' Capabilities to Generate Persuasive Language25 Jun 2024
- 2024A Complete Survey on LLM-based AI Chatbots17 Jun 2024
- 2024An Empirical Design Justice Approach to Identifying Ethical Considerations in the Intersection of Large Language Models and Social Robotics10 Jun 2024
- 2024A Full-duplex Speech Dialogue Scheme Based On Large Language Models29 May 2024
- 2024Building Better AI Agents: A Provocation on the Utilisation of Persona in LLM-based Conversational Agents26 May 2024
50 newest of 172 All 172 (JSON)
- 2024Evaluating Very Long-Term Conversational Memory of LLM Agentscited by 800
- 2022ProsocialDialog: A Prosocial Backbone for Conversational Agentscited by 109
- 2024Building Better AI Agents: A Provocation on the Utilisation of Persona in LLM-based Conversational Agentscited by 26
- 2023Automated Interactive Domain-Specific Conversational Agents that Understand Human Dialogscited by 12
- 2024Self-Emotion Blended Dialogue Generation in Social Simulation Agentscited by 7
- 2026MASCOT: Multi-Agent Socio-Collaborative Companion Systemscited by 7
- 2024Understanding Communication Preferences of Information Workers in Engagement with Text-Based Conversational Agentscited by 6
- 2024Apollonion: Profile-centric Dialog Agentcited by 4
- 2025Toward Real-World Chinese Psychological Support Dialogues: CPsDD Dataset and a Co-Evolving Multi-Agent Systemcited by 2
- 2023Cognitively Inspired Components for Social Conversational Agentscited by 1
- 2026Adaptive Turn-Taking for Real-time Multi-Party Voice Agentscited by 1
- 2025Investigating AI in Peer Support via Multi-Module System-Driven Embodied Conversational Agentscited by 0
- 2026Adaptive Turn-Taking for Real-time Multi-Party Voice Agents11 Jun 2026
- 2026MASCOT: Multi-Agent Socio-Collaborative Companion Systems20 Jan 2026
- 2025Investigating AI in Peer Support via Multi-Module System-Driven Embodied Conversational Agents27 Nov 2025
- 2025Toward Real-World Chinese Psychological Support Dialogues: CPsDD Dataset and a Co-Evolving Multi-Agent System10 Jul 2025
- 2024Understanding Communication Preferences of Information Workers in Engagement with Text-Based Conversational Agents27 Oct 2024
- 2024Self-Emotion Blended Dialogue Generation in Social Simulation Agents3 Aug 2024
- 2024Building Better AI Agents: A Provocation on the Utilisation of Persona in LLM-based Conversational Agents26 May 2024
- 2024Apollonion: Profile-centric Dialog Agent10 Apr 2024
- 2024Evaluating Very Long-Term Conversational Memory of LLM Agents27 Feb 2024
- 2023Cognitively Inspired Components for Social Conversational Agents9 Nov 2023
- 2023Automated Interactive Domain-Specific Conversational Agents that Understand Human Dialogs15 Mar 2023
- 2022ProsocialDialog: A Prosocial Backbone for Conversational Agents25 May 2022
- 2022A Mixture-of-Expert Approach to RL-based Dialogue Managementcited by 16
- 2022A Mixture-of-Expert Approach to RL-based Dialogue Management31 May 2022
- 2021Towards Emotional Support Dialog Systemscited by 294
- 2023A Comprehensive Survey of AI-Generated Content (AIGC): A History of Generative AI from GAN to ChatGPTcited by 221
- 2022BlenderBot 3: a deployed conversational agent that continually learns to responsibly engagecited by 197
- 2022Pretrained Language Models for Text Generation: A Surveycited by 121
- 2023A Complete Survey on Generative AI (AIGC): Is ChatGPT from GPT-4 to GPT-5 All You Need?cited by 84
- 2021DialogueCRN: Contextual Reasoning Networks for Emotion Recognition in Conversationscited by 71
- 2024A Complete Survey on LLM-based AI Chatbotscited by 71
- 2022Control Globally, Understand Locally: A Global-to-Local Hierarchical Graph Network for Emotional Support Conversationcited by 56
- 2023Is ChatGPT Equipped with Emotional Dialogue Capabilities?cited by 54
- 2023Leveraging Large Language Models to Power Chatbots for Collecting User Self-Reported Datacited by 48
- 2021Biomedical Question Answering: A Survey of Approaches and Challengescited by 46
- 2023Faithful Persona-based Conversational Dataset Generation with Large Language Modelscited by 45
12 most cited of 151 All 151 (JSON)
- 2026Neither Silence nor Overlap Is Failure: Intent-Conditioned Evaluation of Turn-Taking in Full-Duplex Spoken Dialogue Models23 Sep 2026
- 2026DeepSAGE: Stage-Aware Reinforcement Learning for Structured CBT Counseling Dialogue23 Aug 2026
- 2026S2Dialog: Multimodal Dialogue Retrieval with Semantic and Acoustic-Style Modeling14 Aug 2026
- 2026I wanted it to feel more personal: Customization of social AI as AI individualism in practice20 Jul 2026
- 2026AuEmoChat: Authentic Emotion Understanding and Rendering for Conversational Speech Synthesis17 Jul 2026
- 2026SURE: Synergistic Uncertainty-aware Reasoning for Multimodal Emotion Recognition in Conversations2 Apr 2026
- 2026Dynamic Fusion-Aware Graph Convolutional Neural Network for Multimodal Emotion Recognition in Conversations22 Mar 2026
- 2026CharacterFlywheel: Scaling Iterative Improvement of Engaging and Steerable LLMs in Production2 Mar 2026
- 2026Hear You in Silence: Designing for Active Listening in Human Interaction with Conversational Agents Using Context-Aware Pacing5 Feb 2026
- 2026Privacy in Human-AI Romantic Relationships: Concerns, Boundaries, and Agency23 Jan 2026
- 2026Scaling Ambiguity: Augmenting Human Annotation in Speech Emotion Recognition with Audio-Language Models21 Jan 2026
- 2026Do You Understand How I Feel?: Towards Verified Empathy in Therapy Chatbots13 Jan 2026
12 newest of 151 All 151 (JSON)