Pith. sign in

REVIEW 4 major objections 6 minor 39 references

MERCI: Multimodal Emotional and peRsonal Conversational Interactions Dataset

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper introduces MERCI, a multimodal dataset of genuine human-robot conversations from 30 participants, together with PERCY, a GPT-4-driven robot that uses facial expressions and sentiment to personalize empathetic responses.

desk verdict MERCI is a useful new HRI dataset that needs its emotion-label validation sorted out before the central claim can be trusted. read the letter →

arxiv 2412.04908 v2 pith:BLDZMPLD submitted 2024-12-06 cs.HC cs.ETcs.RO

Mohammed Althubyani , Zhijin Meng , Shengyuan Xie , Cha Seung , Imran Razzak , Eduardo B. Sandoval , Baki Kocaballi , Francisco Cruz This is my paper · ORCID
classification cs.HCcs.ETcs.RO
keywords human-robotinteractionmultimodaldatasetpersonalizeddialogueempatheticconversationalagentsfacialemotionrecognitionsentimentanalysisGPT-4socialrobotics
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper is trying to establish that real, open-ended human-robot conversations can be captured together with both personal profiles and emotional signals, and that a GPT-4-driven robot can use those signals to hold natural, empathetic dialogue. The contribution is MERCI, a multimodal dataset recorded from 30 participants talking with the ARI robot run by a system called PERCY. PERCY hears the user with Whisper, reads the face with MobileNetV2, scores sentiment with VADER, and feeds all of that plus the participant's questionnaire profile into GPT-4 to generate the next spoken response. Automatic checks and participant questionnaires rate the resulting dialogue high on naturalness, engagement, fluency, relevance, consistency, and empathy. If those ratings hold, MERCI gives the community a resource for studying and building personalized, emotionally aware human-robot interaction.

What carries the argument

The machinery is the PERCY pipeline plus the MERCI corpus it produces. PERCY is a five-module loop on the ARI robot: Whisper converts speech to text; the text goes to both a VADER sentiment analyzer and the GPT-4 response generator; a MobileNetV2-based model classifies the participant's facial expression from the robot's head camera; GPT-4 receives the transcript, sentiment label, emotion label, and the participant's questionnaire profile and generates the next response; the robot speaks it through text-to-speech. That loop creates the dataset's two distinctive ingredients—personal information and emotion state—aligned in time with the conversation. MERCI packages the result as CSV and JSON files containing profiles, chat history, sentiment and emotion labels, and timestamps, together with raw video from the robot head camera and two external cameras, plus a single audio channel.

What would settle it

Take a random sample of roughly 100 utterances from MERCI, have several independent annotators watch the video and audio and choose the same emotion categories the pipeline uses, and measure agreement with the dataset's automated labels; if agreement is near chance levels, the claim that MERCI contains genuine, accurate emotional annotation is falsified.

Watch

Extended reading notes

Core claim

The paper's central claim is that a robot can sustain personalized, emotionally aware conversation using only real-time signals and a pre-collected profile, and that the recordings of those conversations form a reusable multimodal dataset. 'Genuine interactions' matter to the authors: participants volunteered personal information and displayed real emotions, unlike scripted dialogues or TV-derived corpora. The dataset contains 30 participants, an average of 20.27 minutes and 62 utterances per conversation (1,860 utterances total), and 900 emotion labels, with neutral at 53.7 percent and happy at 26.9 percent, plus three video channels and one audio channel per session. Evaluations reported in the paper put automatic readability at perplexity 24.53, cosine similarity 0.32, and MAUVE 0.98, and both GPT-4 and user ratings near 4.5 to 4.7 on a five-point scale across naturalness, engagement, fluency, relevance, consistency, and empathy.

Load-bearing premise

The paper assumes that the automatic face and sentiment classifiers correctly label how participants felt, but reports no accuracy check or human verification of those labels, and if the labels are wrong, the dataset's main novelty as an emotionally annotated resource is not established.

Editorial extensions

If this is right

  • If these results hold, MERCI gives conversational AI researchers a training and evaluation resource built on real human-robot dialogue rather than scripted or TV-derived exchanges.
  • The synchronized video, audio, text, profiles, and emotion labels enable multimodal models that connect what people say with how they look and sound while saying it.
  • PERCY's modular design—Whisper for speech, VADER for sentiment, MobileNetV2 for facial emotion, GPT-4 for generation, and the robot's text-to-speech—offers a reproducible architecture for emotionally aware social robots.
  • The reported numbers (perplexity 24.53, cosine similarity 0.32, MAUVE 0.98, and ratings near 4.5–4.7) give future personalized dialogue systems a concrete baseline to beat.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper: a control condition without the emotion and profile modules would tell whether the high ratings come from GPT-4's general conversational skill or from the personalization itself.
  • Beyond the paper: the participant sample is young, mostly students, and mostly Asian, so the comfort and empathy results may not transfer to older or more diverse users without further data collection.
  • Beyond the paper: because the emotion labels come from automated tools with no human verification, the dataset's value as ground truth for emotion research depends on a future validation study; until then, the labels are best read as system outputs rather than ground truth.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper introduces MERCI, a multimodal dataset of human-robot conversations collected from 30 participants interacting with the ARI robot through the PERCY system. PERCY combines speech-to-text (Whisper), GPT-4-based response generation conditioned on participant profiles and real-time emotion estimates, and text-to-speech. The dataset includes video, audio, transcripts, questionnaire profiles, sentiment labels, and emotion labels. The authors report automatic and user evaluations claiming high naturalness, engagement, fluency, consistency, relevance, and empathy. The paper argues this addresses a gap in open-domain HRI dialogue datasets with genuine interactions and emotional/personal annotations.

Significance. If validated, MERCI would provide a useful resource: real human-robot interaction data with personal profiles, multimodal recordings, and emotional annotations. The user study with 30 participants and standard HRI instruments (NASA TLX, NARS) is a concrete strength, and the promise of public code/data enables reproducibility. However, the central novelty—the emotional annotation—currently rests on unvalidated automatic modules, and the quality claims are supported by evaluations that lack baselines or a fully described protocol. The raw recordings, transcripts, and profiles likely retain value even if the emotional labels need re-annotation or validation.

major comments (4)
  1. [III.D and III.C] The emotion labels—a key claimed contribution—are produced by a MobileNetV2-based facial expression recognizer and VADER sentiment analysis with no reported validation accuracy, training data, or human verification. VADER outputs valence (positive/negative/neutral) rather than the discrete emotion categories shown in Figure 5, and the mapping between the two is unspecified. Since GPT-4 responses are conditioned on these labels, the reliability of the emotion annotations directly affects the quality of the entire dataset. Please provide validation evidence, such as agreement with human annotators on a subset, benchmark accuracy of the emotion recognition model, or a description of the model's training data and source.
  2. [V.B] The emotion distribution is internally inconsistent. The counts listed (489 + 245 + 110 + 44 + 21 + 2 = 911) do not sum to the stated total of 900 emotions, and the percentages in Figure 5 (e.g., 53.7% of 900 = 483.3, not 489) do not match the text counts. Please correct these numbers and present exact counts in a table, as this is a direct factual issue in the dataset description.
  3. [V.D] The automatic evaluation metrics (perplexity 24.53, cosine similarity 0.32, MAUVE 0.98) are reported without baselines or comparisons to other dialogue systems, making them uninterpretable as evidence of 'high quality.' In particular, MAUVE requires an explicit reference distribution of human-like responses, which the paper does not specify. Please provide baseline comparisons (e.g., against other open-domain dialogue models) or clarify the reference distributions used.
  4. [V.D.2] Using GPT-4 to evaluate responses generated by the same GPT-4-based system raises a self-referential concern. The user study offers partial independent grounding, but the GPT-4 evaluation protocol—prompt text, number of samples evaluated, and how scores were aggregated—is not described, and no quantitative agreement with user ratings is reported. Please provide the evaluation details and an inter-rater or correlation analysis between GPT-4 and user evaluations.
minor comments (6)
  1. [Abstract and I] 'To add a systemic gap' should read 'To address a systemic gap.'
  2. [IV.B.4] In the example dialogue, 'netural' is a typo for 'neutral.'
  3. [Figure 3] The captions read 'Video Chanel' and should be 'Video Channel.'
  4. [II.B] 'Multi-model information' should be 'multimodal information' in the description of MELD.
  5. [III.D] The MobileNetV2 backbone is not cited; please add a proper reference for the architecture.
  6. [V.B] Calling a 30-participant, 1860-utterance dataset 'large-scale' is an overstatement relative to existing dialogue corpora (e.g., DailyDialog, MELD); consider qualifying the claim or comparing with HRI-specific datasets.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found: the dataset and PERCY evaluation are empirical and externally grounded by the user study.

full rationale

The paper's central contributions are the recorded MERCI dataset and the PERCY conversational system, built from a questionnaire, a GPT-4-driven robot dialogue conditioned on profile and emotion-module outputs, and subsequent automatic and user evaluations. The claimed chain is empirical rather than mathematical: no equation is fitted to a target quantity and then renamed as a prediction, and the emotion labels are produced by off-the-shelf modules (MobileNetV2 and VADER) rather than defined in terms of the downstream evaluation scores. The lack of reported validation for those modules is a correctness or reproducibility concern, not a circular reduction. The GPT-4-based evaluation does use the same model family that generated the responses, which weakens its independence, but the paper supplements it with a user study that is external to the generation pipeline, and the evaluation is presented as a subjective quality assessment rather than as a first-principles derivation. No load-bearing self-citation appears; the cited GPT-4 evaluation scheme in reference [39] is not authored by the present paper's authors. Consequently, no step in the paper reduces to its own inputs by construction, so the appropriate circularity score is zero.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The central contribution is empirical (a dataset), so the ledger is dominated by assumptions about the accuracy and validity of the measurement and evaluation tools rather than fitted parameters. No free parameters are fitted to the final dataset; the emotion model and prompt are taken from prior work without retraining on this dataset.

assumptions (4)
  • domain assumption Facial expression recognition (MobileNetV2-based) and VADER sentiment analysis provide accurate real-time estimates of participant emotional state.
    Invoked in Sections III.C and III.D; the emotion labels in the dataset depend on these modules, but no accuracy or validation results are reported.
  • domain assumption GPT-4 generates responses that are fluent, relevant, consistent, and empathetic when given a user profile and emotion state.
    The entire system relies on this (Section III.B); the evaluation claims assume this capability.
  • domain assumption Participants' self-reported answers to the initial questionnaire accurately represent their stable personal attributes.
    The personalization pipeline uses these profiles as ground truth (Section IV.A.1).
  • domain assumption The evaluation metrics (perplexity, cosine similarity, MAUVE, GPT-4 Likert ratings) are meaningful indicators of dialogue quality without reference to a human baseline.
    Used in Section V.D as quality evidence; the paper does not establish their validity for HRI dialogue.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MERCI: Multimodal Emotional and peRsonal Conversational Interactions Dataset." pith.science (2026). https://pith.science/paper/BLDZMPLD

@misc{pith2026241204908,
  author       = {Pith},
  title        = {Pith review of: MERCI: Multimodal Emotional and peRsonal Conversational Interactions Dataset},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BLDZMPLD}},
  note         = {Machine review of arXiv:2412.04908}
}
read the original abstract

The integration of conversational agents into our daily lives has become increasingly common, yet many of these agents cannot engage in deep interactions with humans. Despite this, there is a noticeable shortage of datasets that capture multimodal information from human-robot interaction dialogues. To address this gap, we have recorded a novel multimodal dataset (MERCI) that encompasses rich embodied interaction data. The process involved asking participants to complete a questionnaire and gathering their profiles on ten topics, such as hobbies and favorite music. Subsequently, we initiated conversations between the robot and the participants, leveraging GPT-4 to generate contextually appropriate responses based on the participant's profile and emotional state, as determined by facial expression recognition and sentiment analysis. Automatic and user evaluations were conducted to assess the overall quality of the collected data. The results of both evaluations indicated a high level of naturalness, engagement, fluency, consistency, and relevance in the conversation, as well as the robot's ability to provide empathetic responses. It is worth noting that the dataset is derived from genuine interactions with the robot, involving participants who provided personal information and conveyed actual emotions.

Figures

Figures reproduced from arXiv: 2412.04908 by the authors.

Figure 1
Figure 1. PERCY analyzes users’ emotional state through real [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. In addition to the participant and the robot, two external [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The stream data from the ARI robot head camera (a), [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: The participant pool is diverse in terms of age, cultural [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Emotions distribution across all the conversations. [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 29 canonical work pages

  1. [1]

    What is the role of the next generation of cognitive robotics?

    S. Shimoda, L. Jamone, D. Ognibene, T. Nagai, A. Sciutti, A. Costa- Garcia, Y . Oseki, and T. Taniguchi, “What is the role of the next generation of cognitive robotics?” Advanced Robotics , vol. 36, no. 1- 2, pp. 3–16, 2022

    work page 2022

  2. [2]

    FurChat: An Embodied Conversational Agent using LLMs, Combining Open and Closed-Domain Dialogue with Facial Expressions

    N. Cherakara, F. Varghese, S. Shabana, N. Nelson, A. Karukayil, R. Kulothungan, M. A. Farhan, B. Nesset, M. Moujahid, T. Dinkar et al., “Furchat: an embodied conversational agent using llms, combining open and closed-domain dialogue with facial expressions,” arXiv preprint arXiv:2308.15214, 2023

    work page Pith review arXiv 2023

  3. [3]

    A comprehensive study of chatgpt: advancements, limitations, and ethical considerations in natural language processing and cybersecurity,

    M. Alawida, S. Mejri, A. Mehmood, B. Chikhaoui, and O. Isaac Abio- dun, “A comprehensive study of chatgpt: advancements, limitations, and ethical considerations in natural language processing and cybersecurity,” Information, vol. 14, no. 8, p. 462, 2023

    work page 2023

  4. [4]

    Opentab: Advancing large language models as open-domain table reasoners,

    K. Kong, J. Zhang, Z. Shen, B. Srinivasan, C. Lei, C. Faloutsos, H. Rang- wala, and G. Karypis, “Opentab: Advancing large language models as open-domain table reasoners,” arXiv preprint arXiv:2402.14361 , 2024

    arXiv 2024

  5. [5]

    Using ARI: Unboxing and getting started,

    PAL Robotics, “Using ARI: Unboxing and getting started,” https: //pal-robotics.com/blog/using-ari-unboxing-and-getting-started/, 2022, accessed: 2024-09-26

    work page 2022

  6. [6]

    M. F. McTear, Z. Callejas, and D. Griol, The conversational interface . Springer, 2016, vol. 6, no. 94

    work page 2016

  7. [7]

    Conversational agents in online organization–stakeholder interactions: a state-of-the-art analysis and im- plications for further research,

    S. Syv ¨anen and C. Valentini, “Conversational agents in online organization–stakeholder interactions: a state-of-the-art analysis and im- plications for further research,” Journal of Communication Management, vol. 24, no. 4, pp. 339–362, 2020

    work page 2020

  8. [8]

    A survey of chatbot systems through a loebner prize competition,

    L. Brade ˇsko and D. Mladeni ´c, “A survey of chatbot systems through a loebner prize competition,” in Proceedings of Slovenian language technologies society eighth conference of language technologies , vol. 2. sn, 2012, pp. 34–37

    work page 2012

Show all 39 references
  1. [9]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” Advances in neural information processing systems , vol. 30, 2017

    2017

  2. [10]

    Large scale multi-actor generative dialog modeling,

    A. Boyd, R. Puri, M. Shoeybi, M. Patwary, and B. Catanzaro, “Large scale multi-actor generative dialog modeling,” arXiv preprint arXiv:2005.06114, 2020

    2005 arXiv

  3. [11]

    You impress me: Dialogue generation via mutual persona perception,

    Q. Liu, Y . Chen, B. Chen, J.-G. Lou, Z. Chen, B. Zhou, and D. Zhang, “You impress me: Dialogue generation via mutual persona perception,” arXiv preprint arXiv:2004.05388 , 2020

    2004 arXiv

  4. [12]

    Topic-driven and knowledge-aware transformer for dialogue emotion detection,

    L. Zhu, G. Pergola, L. Gui, D. Zhou, and Y . He, “Topic-driven and knowledge-aware transformer for dialogue emotion detection,” arXiv preprint arXiv:2106.01071, 2021

    2021 arXiv

  5. [13]

    An evaluation of human conversational preferences in social human-robot interaction,

    C. Sirithunge, A. B. P. Jayasekara, and D. Chandima, “An evaluation of human conversational preferences in social human-robot interaction,” Applied Bionics and Biomechanics , vol. 2021, pp. 1–13, 2021

    2021

  6. [14]

    Empathy in human–robot interaction: Design- ing for social robots,

    S. Park and M. Whang, “Empathy in human–robot interaction: Design- ing for social robots,” International journal of environmental research and public health , vol. 19, no. 3, p. 1889, 2022

    2022

  7. [15]

    Real-time emo- tion generation in human-robot dialogue using large language models,

    C. Mishra, R. Verdonschot, P. Hagoort, and G. Skantze, “Real-time emo- tion generation in human-robot dialogue using large language models,” Frontiers in Robotics and AI , vol. 10, 2023

    2023

  8. [16]

    The jddc corpus: A large-scale multi-turn chinese dialogue dataset for e-commerce customer service,

    M. Chen, R. Liu, L. Shen, S. Yuan, J. Zhou, Y . Wu, X. He, and B. Zhou, “The jddc corpus: A large-scale multi-turn chinese dialogue dataset for e-commerce customer service,” arXiv preprint arXiv:1911.09969, 2019

    1911 arXiv

  9. [17]

    Challenges in building intelligent open- domain dialog systems,

    M. Huang, X. Zhu, and J. Gao, “Challenges in building intelligent open- domain dialog systems,” ACM Transactions on Information Systems (TOIS), vol. 38, no. 3, pp. 1–32, 2020

    2020

  10. [18]

    Dailydia- log: A manually labelled multi-turn dialogue dataset,

    Y . Li, H. Su, X. Shen, W. Li, Z. Cao, and S. Niu, “Dailydia- log: A manually labelled multi-turn dialogue dataset,” arXiv preprint arXiv:1710.03957, 2017

    2017 arXiv

  11. [19]

    Meld: A multimodal multi-party dataset for emotion recognition in conversations,

    S. Poria, D. Hazarika, N. Majumder, G. Naik, E. Cambria, and R. Mihal- cea, “Meld: A multimodal multi-party dataset for emotion recognition in conversations,” arXiv preprint arXiv:1810.02508 , 2018

    2018 arXiv

  12. [20]

    Will i sound like me? improving persona consistency in dialogues through pragmatic self-consciousness,

    H. Kim, B. Kim, and G. Kim, “Will i sound like me? improving persona consistency in dialogues through pragmatic self-consciousness,” arXiv preprint arXiv:2004.05816, 2020

    2004 arXiv

  13. [21]

    A survey on empathetic dialogue systems,

    Y . Ma, K. L. Nguyen, F. Z. Xing, and E. Cambria, “A survey on empathetic dialogue systems,” Information Fusion , vol. 64, pp. 50–70, 2020

    2020

  14. [22]

    Whisper: Openai’s speech recognition model,

    OpenAI, “Whisper: Openai’s speech recognition model,” https://openai. com/index/whisper/, 2023, accessed: 2024-09-27

    2023

  15. [23]

    Natural language toolkit,

    S. Bird, E. Klein, and E. Loper, “Natural language toolkit,” 2023, accessed: 2024-09-27. [Online]. Available: https://www.nltk.org/

    2023

  16. [24]

    Ari sdk documentation - speech module,

    PAL Robotics, “Ari sdk documentation - speech module,” 2023, accessed: 2024-09-27. [Online]. Available: https://docs.pal-robotics. com/ari/sdk/23.1.15/speech.html

    2023

  17. [25]

    Factors for personalization and localization to optimize human–robot interaction: A literature review,

    N. Gasteiger, M. Hellou, and H. S. Ahn, “Factors for personalization and localization to optimize human–robot interaction: A literature review,” International Journal of Social Robotics , vol. 15, no. 4, pp. 689–701, 2023

    2023

  18. [26]

    The study of speech naturalness in communication disorders: A systematic review of the literature,

    M. Klopfenstein, K. Bernard, and C. Heyman, “The study of speech naturalness in communication disorders: A systematic review of the literature,” Clinical linguistics & phonetics , vol. 34, no. 4, pp. 327–338, 2020

    2020

  19. [27]

    Evaluation of the naturalness of chatbot applications,

    A. Atiyah, S. Jusoh, and F. Alghanim, “Evaluation of the naturalness of chatbot applications,” in 2019 IEEE Jordan International Joint Con- ference on Electrical Engineering and Information Technology (JEEIT) . IEEE, 2019, pp. 359–365

    2019

  20. [28]

    Towards a method for evaluating naturalness in conversational dialog systems,

    V . Hung, M. Elvir, A. Gonzalez, and R. DeMara, “Towards a method for evaluating naturalness in conversational dialog systems,” in 2009 IEEE international conference on systems, man and cybernetics. IEEE, 2009, pp. 1236–1241

    2009

  21. [29]

    Usefulness, localizability, humanness, and language-benefit: additional evaluation criteria for natural language dialogue systems,

    B. AbuShawar and E. Atwell, “Usefulness, localizability, humanness, and language-benefit: additional evaluation criteria for natural language dialogue systems,” International Journal of Speech Technology, vol. 19, pp. 373–383, 2016

    2016

  22. [30]

    A wizard- of-oz study on a non-task-oriented dialog systems that reacts to user engagement,

    Z. Yu, L. Nicolich-Henkin, A. W. Black, and A. Rudnicky, “A wizard- of-oz study on a non-task-oriented dialog systems that reacts to user engagement,” in Proceedings of the 17th annual meeting of the Special Interest Group on Discourse and Dialogue , 2016, pp. 55–63

    2016

  23. [31]

    Fluency in monologic and dialogic task performance: Challenges in defining and measuring l2 fluency,

    P. Tavakoli, “Fluency in monologic and dialogic task performance: Challenges in defining and measuring l2 fluency,” International Review of Applied Linguistics in Language Teaching , vol. 54, no. 2, pp. 133– 150, 2016

    2016

  24. [32]

    A retrieval-based dialogue system utilizing utterance and context embeddings,

    A. Bartl and G. Spanakis, “A retrieval-based dialogue system utilizing utterance and context embeddings,” in 2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA) . IEEE, 2017, pp. 1120–1125

    2017

  25. [33]

    Transformer-based empathetic response generation using dialogue situation and advanced- level definition of empathy,

    Y .-H. Wang, J.-H. Hsu, C.-H. Wu, and T.-H. Yang, “Transformer-based empathetic response generation using dialogue situation and advanced- level definition of empathy,” in 2021 12th International Symposium on Chinese Spoken Language Processing (ISCSLP) . IEEE, 2021, pp. 1–5

    2021

  26. [34]

    Consistent dialogue across concurrent topics based on an expert system model,

    B.-s. Lin, H.-m. Wang, and L.-s. Lee, “Consistent dialogue across concurrent topics based on an expert system model,” in Sixth European Conference on Speech Communication and Technology , 1999

    1999

  27. [35]

    A novel two-stage generation framework for promoting the persona-consistency and diversity of responses in neural dialog systems,

    T. Shi and Y . Song, “A novel two-stage generation framework for promoting the persona-consistency and diversity of responses in neural dialog systems,” IEEE Transactions on Neural Networks and Learning Systems, 2021

    2021

  28. [36]

    Language model evaluation beyond per- plexity,

    C. Meister and R. Cotterell, “Language model evaluation beyond per- plexity,” arXiv preprint arXiv:2106.00085 , 2021

    2021 arXiv

  29. [37]

    Distance weighted cosine similarity measure for text classification,

    B. Li and L. Han, “Distance weighted cosine similarity measure for text classification,” in Intelligent Data Engineering and Automated Learning–IDEAL 2013: 14th International Conference, IDEAL 2013, Hefei, China, October 20-23, 2013. Proceedings 14 . Springer, 2013, pp. 611–618

    2013

  30. [38]

    Mauve: Measuring the gap between neural text and human text using divergence frontiers,

    K. Pillutla, S. Swayamdipta, R. Zellers, J. Thickstun, S. Welleck, Y . Choi, and Z. Harchaoui, “Mauve: Measuring the gap between neural text and human text using divergence frontiers,” Advances in Neural Information Processing Systems , vol. 34, pp. 4816–4828, 2021

    2021

  31. [39]

    Gpt-4v (ision) is a human-aligned evaluator for text-to-3d generation,

    T. Wu, G. Yang, Z. Li, K. Zhang, Z. Liu, L. Guibas, D. Lin, and G. Wetzstein, “Gpt-4v (ision) is a human-aligned evaluator for text-to-3d generation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 22 227–22 238

    2024

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.