From Eliza to XiaoIce: challenges and opportunities with social chatbots
Open accessThe DOI matches its Crossref record; 81 of the 94 references it cites were found in a registry or an index.
Abstract
Conversational systems have come a long way since their inception in the 1960s. After decades of research and development, we have seen progress from Eliza and Parry in the 1960s and 1970s, to task-completion systems as in the Defense Advanced Research Projects Agency (DARPA) communicator program in the 2000s, to intelligent personal assistants such as Siri, in the 2010s, to today’s social chatbots like XiaoIce. Social chatbots’ appeal lies not only in their ability to respond to users’ diverse requests, but also in being able to establish an emotional connection with users. The latter is done by satisfying users’ need for communication, affection, as well as social belonging. To further the advancement and adoption of social chatbots, their design must focus on user engagement and take both intellectual quotient (IQ) and emotional quotient (EQ) into account. Users should want to engage with a social chatbot; as such, we define the success metric for social chatbots as conversation-turns per session (CPS). Using XiaoIce as an illustrative example, we discuss key technologies in building social chatbots from core chat to visual awareness to skills. We also show how XiaoIce can dynamically recognize emotion and engage the user throughout long conversations with appropriate interpersonal responses. As we become the first generation of humans ever living with artificial intelligenc (AI), we have a responsibility to design social chatbots to be both useful and empathetic, so they will become ubiquitous and help society as a whole.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“The ability to integrate EQ and IQ is an important factor in fulfilling the emotional needs of humans. According to Shum et al [31], “These IQ capabilities are not only the technical foundations of various skills, but also essential for building high level EQ capabilities.””
From User Experiences of Social Support From Companion Chatbots in Everyday Contexts: Thematic Analysis (2020) · cites this paper as [31]
In the citing paper
“behaviors that have been historically very difficult to accurately capture in AI, but are essential if AI is to serve as effective companions for humans [31]”
From User Experiences of Social Support From Companion Chatbots in Everyday Contexts: Thematic Analysis (2020) · cites this paper as [31]
In the citing paper
“Studies of social chatbots such as XiaoIce (Shum et al., 2018) and Replika (Ta et al., 2020) and the mental health chatbots Woebot (Fitzpatrick et al., 2017) and Tess (Fulmer et al., 2018) demonstrate that chatbots may induce a sense of a relationship in users.”
From My Chatbot Companion - a Study of Human-Chatbot Relationships (2021) · cites this paper as Shum et al., 2018
In the citing paper
“From the emergence of the first chatbot, ELIZA, developed by Joseph Weizenbaum (1966), chatbots have been trying to ‘mimic human behaviour in a text-based conversation’ (Shum et al. 2018, p. 10; Abd-Alrazaq et al. 2020).”
From Chatbot breakthrough in the 2020s? An ethical reflection on the trend of automated consultations in health care (2022) · cites this paper as Shum et al. 2018
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 2×
“, 2016), computer vision (Krizhevsky et al., 2012; He et al., 2016), information retrieval (Huang et al.”
BackgroundImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 2×
“, 2016), computer vision (Krizhevsky et al., 2012; He et al., 2016), information retrieval (Huang et al.”
BackgroundHochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318 · cited 3×
“These representation vectors are then fed into a decoder, usually another LSTM, to generate the response word-by-word (Vinyals and Le, 2015).”
BackgroundMaslow, A. H. (1943) · Psychological Review · doi:10.1037/h0054346 · Cited by 21,563 · cited 2×
“Unlike early chatbots designed for chitchat, social chatbots are created to serve users’ needs for communication, affection, and social belonging (Maslow, 1943) rather than for passing the Turing Test.”
BackgroundSequence to Sequence Learning with Neural Networks
Sutskever, I., Vinyals, O., & Le, Q. V. (2014) · arXiv · doi:10.48550/arxiv.1409.3215 · Cited by 17,766
Neural Machine Translation by Jointly Learning to Align and Translate
Bahdanau, D., Cho, K., & Bengio, Y. (2014) · arXiv · doi:10.48550/arxiv.1409.0473 · Cited by 16,213 · cited 3×
“…perceptual and cognitive AI technologies, e.g., natural language understanding (Bengio et al., 2001; Mikolov et al., 2013; Sutskever et al., 2014; Bahdanau et al., 2015; Mesnil et al., 2015), speech recognition and synthesis (Hinton et al., 2012; Deng et al., 2013; Xiong et al., 2016; Qian et…”
BackgroundMethodFrames of Mind: The Theory of Multiple Intelligences
Klitgaard, R., & Gardner, H. (1984) · Journal of Policy Analysis and Management · doi:10.2307/3324560 · Cited by 13,636
(2017) · Defining and Classifying Children in Need · doi:10.4324/9781315258324-16 · Cited by 11,320
I.—COMPUTING MACHINERY AND INTELLIGENCE
TURING, A. M. (1950) · Mind · doi:10.1093/mind/lix.236.433 · Cited by 10,290 · cited 4×
“Unlike early chatbots designed for chitchat, social chatbots are created to serve users’ needs for communication, affection, and social belonging (Maslow, 1943) rather than for passing the Turing Test.”
BackgroundMethodInfluentialHinton, G., Deng, L., Yu, D., Dahl, G., Mohamed, A.-R., Jaitly, N., Senior, A., Vanhoucke, V., Nguyen, P., Sainath, T., et al. (2012) · IEEE Signal Processing Magazine · doi:10.1109/msp.2012.2205597 · Cited by 10,260 · cited 2×
“, 2013, 2015), speech recognition and synthesis (Hinton et al., 2012; Deng et al., 2013; Qian et al., 2014; van den Oord et al., 2016; Xiong et al., 2016), computer vision (Krizhevsky et al.”
Background
Research works printed but not linked: 8
- Bengio, 2003, A neural probabilistic language model, 1137
- 10.1016/0004-3702(71)90002-6 · DOI · Paper page
- 10.21437/icslp.2000-224 · DOI
- 10.18653/v1/P16-1094 · DOI · Paper page
- Mikolov, 2013, Distributed representations of words and phrases and their compositionality, 3111
- 10.21437/Eurospeech.1999-389 · doi:10.21437/eurospeech.1999-389
- 10.18653/v1/e17-1042 · DOI · Paper page
- Yu, 2016, Chatbot evaluation and database expansion via crowdsourcing
Cited by
Items 701–746 of 746
Li, L., & Li, Y. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6154908
Silent Signals, Loud Consequences: Chemo-Communication, Ethics, and Digital Therapeutics
Brasi, C., & Seccomandi, B. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-14160-6_3
Generative AI and Education: Student Voices, Ethical Challenges, and Classroom Innovations
Moura, A. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-2600-0101-1.ch003
Synthetica: The First AI-Native Cohort and the Democratic Risks of Synthetic Infrastructures
Milelli, R. (2026) · Journal of Artificial Intelligence & Cloud Computing · doi:10.47363/jaicc/2026(5)515
Agarwal, C., & Krishnaswamy, V. (2026) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2026.2639098
An Experimental Study on Harassment Moderation in Llama and Alpaca
de Sousa, H. T., & Paschoal, L. N. (2026) · Big Data and Cognitive Computing · doi:10.3390/bdcc10040100
“…each type of system reacted distinctly: commercial systems (like Alexa [7] and Siri [8]) tended to be evasive; rule-based chatbots (like E.L.I.Z.A [29] and A.L.I.C.E. [ 30]) often deflected the topic; and data-driven systems (like Cleverbot [31]) presented a risk of generating responses that…”
AI-Powered Petition Chatbot for Smart Grievance Management and Urgency Analysis
Parameswari, M., Mohanraj, V., Jayasuriya, B., Jeeva Mary, S., Keerthana, G., & Yuvaprasanth, L. (2026) · International Conference on Communication, Computing and Electronics Systems (ICCCES) · doi:10.1109/iccces62661.2026.11436618
“Similarly, Shum et al. reviewed the evolution from early rule-based systems like Eliza to advanced social chatbots such as Xiaolce, emphasizing the importance of natural dialogue flow, emotion handling, and multi-turn conversation capabilities [7].”
Lee, J.-C., & Huang, X. (2026) · Industrial Management & Data Systems · doi:10.1108/imds-04-2025-0416
Gao, Y., Piao, Y., & Xuan, D. (2026) · Sustainability · doi:10.3390/su18073554 · Cited by 2
AVATAR-MUSE: An Automated, Avatar-Based Chatbot System for Museum Ticket Booking
Jha, P., Kumar, S. P., Manoj, S., Sinduja, A., & Murugan, S. (2026) · International Conference on Multi-Agent Systems for Collaborative Intelligence (ICMSCI) · doi:10.1109/icmsci67830.2026.11469688
“As tourists and local visitors alike seek more personalized and convenient services, the demand for digital transformation within museum operations especially ticketing and information systems has grown exponentially[2][3].”
Artificial Intelligence as a Catalyst for Curiosity in Teaching, Learning, and Education
Garner, B., & Snyder, T. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/407439
Yang, Y., Tavares, J., & Oliveira, T. (2026) · International Journal of Information Management Data Insights · doi:10.1016/j.jjimei.2026.100413 · Cited by 1
Oishi, N. N., & Chowdhury, A. S. (2025) · International Conference on Computer and Information Technology (ICCIT) · doi:10.1109/iccit68739.2025.11490177
“T. Young and D. Hazarika [4] highlighted the relevance of attention processes, and Shum et al. [5] proved the role of emotional alignment and long-term engagement in social chatbots like XiaoIce.”
Zhu, Y. (2025) · International Conference on Information Education and Artificial Intelligence (ICIEAI) · doi:10.1145/3799457.3799660
“Early chatbots relied primarily on keyword recognition to generate human-like responses and provide context-based feedback [10].”
Abdalla, J., Naser, M. Z., & Alogla, S. M. (2026) · Emerald Group Publishing Limited eBooks · doi:10.1108/978-1-83662-866-820261013
Chen, Y., Sun, Y., Zhang, Y., & Wang, Y. (2026) · Information Technology and People · doi:10.1108/itp-03-2025-0362
Jin, Y., & Chen, Y. (2026) · Virtual Reality · doi:10.1007/s10055-026-01383-6
“This evolution has enabled CAs to move beyond handling simple task-oriented commands, gradually transforming into intelli - gent agents with social functionalities (Shum et al. 2018 ).”
Bridging the silos in affective AI: a critical perspective from data to society
Inoshita, K. (2026) · AI & Society · doi:10.1007/s00146-026-03324-y · Cited by 3
Leszczyński, G., Gaczek, P., & Mouakher, A. (2026) · Journal of Research in Interactive Marketing · doi:10.1108/jrim-09-2025-0531
Insights and future trends of humanised AI in organisations
Nalvaiz, N. L., & Alesón, M. L. R. (2026) · International journal of organizational analysis · doi:10.1108/ijoa-05-2025-5496
Role of AI chatbots in nutrition and dietary guidance
Agarwal, S., & Singh, V. (2026) · Elsevier eBooks · doi:10.1016/b978-0-443-40501-3.00014-5
Zhang, Y., & Zhou, Z. (2026) · Bulletin of Counseling and Psychotherapy · doi:10.51214/002026081913000
“Social chatbot research has long argued that dialogue systems face a trade-off among scale, controllability, and social appropriateness (Shum et al., 2018).”
Sou, K. L., Yow, W. Q., Lau, M. Z. Y., & Ouyang, F. (2026) · Innovation in Aging · doi:10.1093/geroni/igag073 · Cited by 1
“Shum et al. (2018) categorize chatbots into three types: task-completion, personal-assistant, and social.”
Vora, A., Parpia, S., & Gada, T. N. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-30836-8_22
‘She lifted me up’: kinship construction in human-AI resilient communication among Chinese users
Li, Y. W., Luo, M., & Ren, H. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1832594
“These systems engage users in sustained, empathetic everyday dialogue, provide social support, and facilitate relationship formation (Ho et al., 2018; Shum et al., 2018; Zhou et al., 2020).”
How Identity Types and Intelligence Levels of Digital Humans Affect Consumer Purchase Decisions
Qi, T., Wang, T., & Tang, J. (2026) · International Journal of Electronic Commerce · doi:10.1080/10864415.2026.2686040
“Unlike low-intelligence versions that rely on template-based matching, high-intelligence digital humans can engage in fluent, contextually adaptive conversations [94].”
Turgay, A., Sancakdaroglu, H., Çalışır, F., & Cetinguc, B. (2026) · Lecture notes in management and industrial engineering · doi:10.1007/978-3-032-24958-6_14
Ashrafi, N., Kupkovic, R., Vona, F., Hinzmann, S., Vergari, M., Wang, Y., Braytee, A., Gross, I., Pasoglou, J., & Voigt‐Antons, J. (2026) · International Conference on Quality of Multimedia Experience (QoMEX) · doi:10.1109/qomex69967.2026.11618347
“Prior work has typically focused either on conversational improvements with low-fidelity avatars [7], [8] or on realistic visuals without real-time interaction [9], leaving the combined effects of realism and emotional expressiveness underexplored.”
Understanding Virtual Assistants: A Systematic Review
Montiel, C. S. A., Rodríguez-García, J. G., & Mendoza, S. (2026) · Information Systems Frontiers · doi:10.1007/s10796-026-10788-2
Yin, J., Wang, H., Bao, P., Xu, J., & Wang, Y. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770854.3783953 · Cited by 15
Liu, S., Zheng, C., & Zeng, N. (2026) · Information & Management · doi:10.1016/j.im.2026.104412
Sánchez-Suárez, G., & Quiñones, L. M. T. (2026) · Respuestas · doi:10.22463/0122820x.5760
Wheadon, J., Hayden‐Evans, M., Hryniewicki, J., Afsharnejad, B., Tan, T., Bölte, S., & Girdler, S. J. (2026) · Social Sciences & Humanities Open · doi:10.1016/j.ssaho.2026.103497
Liu, D., Pan, L., Sabri, A. Q. M., & Fan, G. F. (2026) · ACM Transactions on Social Computing · doi:10.1145/3844666 · Cited by 1
Perry, A., & Cameron, C. D. (2026) · Cambridge University Press eBooks · doi:10.1017/9781009639057.002
Models Match or Surpass Objective Human Performance on Various Tasks
Goldstein, A., & Stanovsky, G. (2026) · Cambridge University Press eBooks · doi:10.1017/9781009639057.006
Cantwell, S., & Goldkind, L. (2026) · Oxford University Press eBooks · doi:10.1093/9780197851364.003.1706
Piispanen, J.-R., Myllyviita, T., Vakkuri, V., & Rousi, R. A. (2026) · Information and Software Technology · doi:10.1016/j.infsof.2026.108340
Mishra, D. K., Dhal, S. K., Panda, G., & Dhal, S. (2026) · Natural Resources for Human Health · doi:10.53365/nrfhh.1384 · Cited by 1
The Ontology Driven E-Commerce Chatbot for Fashion Industry
Mairaj, M., Ahmed, S., Hidayat, K., Shaikh, A., & Naqvi, S. (2026) · EMITTER International Journal of Engineering Technology · doi:10.24003/emitter.v14i1.1055
“Chatbots were initially designed as chitchat systems. to speak as a non-directive psychotherapist would in a conversation, Utilizing straightforward pattern matching and a response tem plate, the chatbot ELIZA (26)was developed.”
ArtiQ: An Artist-in-the-Loop Chatbot for Question Answering in Art Galleries
Liu, J., Law, E. L., & Shum, H. P. H. (2026) · Nordic Conference on Human-Computer Interaction · doi:10.1145/3829807.3829831
Self-citationBudiman, A. M., Verrell, C., Kristiawan, H. E., Chowanda, A., & Anderies (2026) · International Conference on ICT for Smart Society (ICISS) · doi:10.1109/iciss71466.2026.11715172
“We propose a hybrid architecture combining intent classification, an important step towards understanding user inputs [11], and a Local Large Language Model (LLM) for generating more natural responses.”
Shi, M., Liang, M., & Hwang, B. (2026) · Journal of Management in Engineering · doi:10.1061/jmenea.meeng-7659
Dhanju, S. S., Panigrahi, G. R., Khomane, R., Jit, H., Saini, I., & Singla, C. (2026) · International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC) · doi:10.1109/assic70037.2026.11716796
Synthetic Intimacy: Relational AI and the Interface of the Self
Shin, D. (2026) · Springer eBooks · doi:10.1007/978-3-032-39278-7_6
Metadata
- Authors
- Shum, H.-Y., He, X.-D., & Li, D.
- Affiliations
- Microsoft (United States) (US)
- Corresponding author
- Shum, Heung-yeung
- Venue
- Frontiers of Information Technology & Electronic Engineering
- Publisher
- Zhejiang University Press
- Published
- 2018-01
- Volume
- 19
- Issue
- 1
- Pages
- 10-26
- ISSN
- 2095-9184 (ISSN-L)
- 2095-9230
- OpenAlex
- W2964309167
- Cited by
- 749
- References
- 94
- Citations by year
- 2026: 60 · 2025: 96 · 2024: 83 · 2023: 118 · 2022: 103 · 2021: 109 · 2020: 103 · 2019: 56 · 2018: 21
- FWCI
- 40.13
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- Frontiers of Information Technology & Electronic Engineering (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Rare & Special e-Zone (The Hong Kong University of Science and Technology) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Microsoft Research
- Microsoft Research Asia
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Speech and dialogue systems
- Social Sciences › Business, Management and Accounting › Marketing › AI in Service Interactions
- Social Sciences › Psychology › Social Psychology › Social Robot Interaction and HRI
- Keywords
- social chatbots, Eliza, emotional connection, user engagement, emotional quotient, emotion recognition, visual awareness, dialogue systems, human-AI interaction
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.