Pith. sign in

REVIEW 3 major objections 6 minor 46 references

DEBATE: A Dataset for Disentangling Textual Ambiguity in Mandarin Through Speech

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The DEBATE dataset pairs 1,001 ambiguous Mandarin utterances with 10,010 recordings from ten native speakers and shows large speech-language models resolve spoken intent far worse than human listeners.

desk verdict A genuinely useful Mandarin speech-ambiguity dataset, with a benchmark that overclaims its control over text-only priors. read the letter →

arxiv 2506.07502 v1 pith:PQDEYPYR submitted 2025-06-09 cs.CL

Haotian Guo , Jing Han , Yongfeng Tu , Shihao Gao , Shengfan Shen , Wulong Xiang , Weihao Gan , Zixing Zhang This is my paper · ORCID
classification cs.CL
keywords speech-baseddisambiguationMandarinChinesespeech-textdatasetprosodypolyphoniccharactersstructuralambiguitystressandintonationlargespeech-languagemodels
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

The paper sets out to fill a gap: many written Chinese sentences are ambiguous—a character with two readings, a missing word boundary, or a shifted stress can change the meaning—and existing corpora pair such sentences only with text, not with speech. DEBATE pairs 1,001 ambiguous written utterances with 10,010 spoken recordings by ten native speakers, annotating the pronunciation, pause, and stress cues that disambiguate each item. The authors then benchmark three large speech-language models and report that none resolves the intended meaning nearly as well as human listeners, with stress-based ambiguity being the hardest. If the dataset and results hold, they give the field a standard evaluation and training resource for disambiguation through speech, with implications for speech understanding and text-to-speech systems.

What carries the argument

The load-bearing object is the DEBATE corpus, organized into three disambiguation scenarios: polyphonic characters, where pronunciation determines meaning; structural ambiguity, where pause boundaries determine sentence segmentation; and focus ambiguity, where stress and intonation determine semantic emphasis. Each item pairs a written sentence that is ambiguous on the page with a spoken recording annotated for the disambiguated meaning and the prosodic cues that carry it. This pairing turns disambiguation through speech into a measurable task: a model hears one audio recording and must choose the intended reading, while the human baseline on the same items quantifies what the acoustic cues actually convey.

What would settle it

Measure, for every stress item and every speaker, the fundamental-frequency peak, duration, and intensity of the intended stressed syllable relative to the same syllable in the alternative reading, and measure the actual silent intervals at the annotated pause boundaries in the pause items. If a substantial fraction of recordings lack a reliable acoustic difference for the annotated cue, the assumption that DEBATE encodes the targeted prosodic cues fails, and the benchmark numbers would need re-scoring before being read as model ability.

Watch

Extended reading notes

Core claim

The paper claims that Mandarin's written ambiguity can be systematically disentangled by spoken cues, and introduces DEBATE as the first dataset built to study this. It curates 1,001 ambiguous sentences—200 polyphonic-character, 401 structural, and 400 focus-related—each read by 10 native speakers, yielding 10,010 recordings totaling 9.66 hours with semantic and prosodic annotations. On this corpus, three large speech-language models reach only about 51–68 percent accuracy, markedly below human listeners on the same items, with the largest deficit in stress-based disambiguation. The authors read this as evidence that current models can detect relatively explicit cues like pauses but not fine prosody such as stress and pitch movement.

Load-bearing premise

The entire benchmark depends on the ten volunteer speakers actually producing the prescribed pronunciations, pause placements, and stress patterns when they read the instructions; if the recorded prosody does not match the annotations, the reported human–model gap could be an artifact of the recordings rather than a measure of disambiguation ability.

Editorial extensions

If this is right

  • Speech-language models have measurable, task-specific weaknesses: pause-based structure is the easiest cue for them, while stress and intonation is the hardest, so future acoustic-modeling work should concentrate on fine prosody.
  • DEBATE can serve as a common benchmark for disambiguation through speech, letting future models be compared against the same human baseline on the same utterances.
  • The corpus offers targeted training data for text-to-speech systems, particularly for pronouncing polyphonic characters correctly and rendering prosodic stress naturally.
  • The implicit regional variation among the ten speakers makes the dataset usable for studying Mandarin pronunciation variants across dialect backgrounds.
  • If the benchmark is reused, stress-based tasks may require new training objectives or prosody-labeled pretraining data before model performance approaches human levels.

Reading between the lines

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

  • Editorial inference: the paper's three-way taxonomy of pronunciation, pause, and stress suggests a difficulty gradient that may generalize to other languages without explicit word boundaries, so analogous datasets could be built for Cantonese, Japanese, or other isolating languages.
  • Editorial inference: an acoustic verification study could test whether the recorded stress cues are perceptually consistent across the ten speakers; if they are not, the stress-task numbers would need to be reinterpreted as a measure of recording prompt-following rather than model ability.
  • Editorial inference: the human–model gap may shrink fastest on pause tasks because the cue is discrete and temporally located, whereas stress tasks may need prosody-aware training signals that are largely absent from current automatically transcribed corpora.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper introduces DEBATE, a Mandarin speech-text dataset for studying how speech cues (pronunciation, pause, stress, intonation) resolve ambiguity that persists in written text. The dataset contains 1,001 ambiguous sentences, each recorded by 10 native speakers, yielding 10,010 audio samples totaling 9.66 hours, organized into three subtasks: polyphonic-character ambiguity, structural/pause ambiguity, and stress/focus ambiguity. The authors describe a data pipeline involving collection from corpora, social media, and exam banks; LLM-based augmentation with manual verification; a two-person recording protocol; and ASR-based quality checks. They benchmark three large speech-language models (Qwen2-Audio, Qwen2.5-Omni, Gemini 2.0 Flash) in a zero-shot, forced-choice setting and report accuracy and macro-F1 per subtask. They additionally compare models with three human volunteers on a 50-sample-per-subtask subset and report, via violin plots, that humans outperform all models, especially on stress-based disambiguation.

Significance. If the dataset is validated as claimed, DEBATE would be a useful and timely resource: to my knowledge it is the first public Mandarin dataset specifically designed for speech-based disambiguation, and the paper documents the collection pipeline in commendable detail. The release of data and code, the use of multiple speakers with demographic diversity, the two-person recording protocol, and the ASR-based quality checks are all concrete strengths that make the resource potentially reusable for training and evaluation. However, the benchmark evidence for the central claim—that current LSLMs fail to leverage acoustic information—is incomplete: there is no text-only control, the human evaluation is reported only as violin plots with no numeric accuracy or agreement statistics, and the recordings' prosodic validity is not directly established beyond thin manual sampling and ASR lexical alignment. These gaps are fixable within the paper's scope, and the dataset contribution itself remains defensible.

major comments (3)
  1. [§4.2, Table 4] The central empirical claim that LSLMs 'still struggle to effectively leverage acoustic information' is underdetermined by the reported experiments. Table 4 reports only audio-plus-prompt accuracies; no matched text-only condition is shown. Without a text-only baseline (for example, feeding the same forced-choice prompt with the transcript in place of audio to a text LLM, or masking the audio), the accuracies of 51–68% could reflect lexical-semantic priors rather than a failure to use speech cues. A model might choose the statistically more frequent reading of an ambiguous sentence regardless of audio, and the audio could even be slightly harmful. Please add matched text-only baselines and report the audio-minus-text accuracy delta; the conclusion about acoustic information should be revised on that basis.
  2. [§4.2, Figure 4] The human–model comparison is not quantitatively documented. Only three volunteers, 50 samples per subtask, are described, with no accuracy numbers, no confidence intervals, and no inter-annotator agreement. The claim of a 'clear and huge' gap cannot be evaluated from a violin plot alone. Please report the numeric human accuracy per subtask and per speaker, the corresponding model accuracies on the same 50-sample set, and an agreement measure such as Fleiss' kappa or pairwise Cohen's kappa. This is load-bearing because the gap between humans and models is a headline result of the paper.
  3. [§3.2.2–3.2.3, Table 1] The dataset's validity rests on the assumption that the recruited speakers reliably produced the prescribed prosodic cues (pronunciation, pause position, stress). The ASR CER results in Table 1 verify lexical alignment but not prosodic realization: stress, in particular, typically does not change the character sequence and thus is invisible to CER. The manual check of ten randomly selected samples per task is thin, and no criteria or reliability statistics are reported. Please add direct validation of the prosodic cues—for example, acoustic measurements of pause durations and pitch/intensity contours, or a perception test in which naive listeners select the intended meaning from the audio. Without this, the human–model gap could reflect recording artifacts rather than inherent model limitations.
minor comments (6)
  1. [Table 4, Figure 4] There are typographical errors: 'TProun' should be 'TPronun' (or a consistent abbreviation), and 'Gemeni' should be 'Gemini'.
  2. [§3.2.1] No inter-annotator agreement is reported for the manual selection and semantic annotation of the ambiguous sentences; even a brief description of annotation guidelines and a kappa value would strengthen confidence in the classification into the three subtask types.
  3. [§3.2.3, Table 1] The CER comparison with AISHELL-1 and AISHELL-2 is not apples-to-apples because DEBATE sentences are deliberately ambiguous and include polyphonic characters; please report CER per speaker and per subtask, and state the ASR decoding configuration used.
  4. [Figure 1, Table 2] The percentages shown in Figure 1 (e.g., '40%', '20%') are not clearly defined; the text reports 200, 401, and 400 samples per subtask, so the figure should use consistent counts or percentages with explicit labels.
  5. [§3.1] The term 'polyphonic character ambiguity' is nonstandard; consider using 'polyphone ambiguity' or 'heteronym ambiguity' and define the term on first use to avoid confusion with musical polyphony.
  6. [§4.1] The inference setup lacks reproducibility details: please report the decoding parameters, number of runs, and whether the reported variance in Table 4 is across speakers only or across repeated inference runs.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper constructs a dataset and runs external benchmarks, with no derivation whose conclusion is encoded in its inputs.

full rationale

DEBATE is an empirical dataset-construction and benchmarking paper. There is no mathematical derivation, fitted parameter, or predictive model whose output is defined in terms of its inputs. The central claims are (1) that the dataset contains ambiguous Mandarin utterances with speech cues, and (2) that three large speech-language models perform worse than human listeners on a forced-choice disambiguation task. Both claims are supported by direct measurement rather than by construction. The quality-control section uses ASR transcription to measure character error rate on the same recordings; this is a sanity check on audio-text alignment, not a prediction, so it does not constitute circularity. The human evaluation uses the same forced-choice format as the model evaluation, which is actually a strength of the comparison rather than a circular step. The absence of a text-only control condition for the models is a substantive experimental-design limitation, but it is a threat to the interpretation of the model results, not a case where the paper's conclusion is equivalent to its assumptions by definition. Self-citations appear only as ordinary references to prior work and are not load-bearing for the dataset's validity or for the benchmark outcome. No circular step can be exhibited from the paper's text, so the appropriate finding is no significant circularity.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper introduces a dataset and benchmark, not a model or theory, so there are no free parameters or invented entities. It relies on domain assumptions about ambiguity categorization, prosodic annotation validity, and the correctness of human-reviewed semantic labels.

assumptions (3)
  • domain assumption The three ambiguity types (polyphonic characters, structural/pause, focus/stress) cover representative text ambiguities resolvable through speech and are mutually exclusive.
    This categorization is used to select and organize all 1,001 dataset items; if the categories are not representative or overlap, coverage claims weaken. Invoked in Section 3.1.
  • domain assumption The prosodic annotations (pronunciation marks, '/' pause markers, stress indicators) correctly specify the disambiguating cues that a native speaker can produce and a listener can recover.
    Speakers were instructed to follow these annotations, and the dataset's value depends on the recordings actually containing those cues. Invoked in Sections 3.2.1 and 3.2.2.
  • domain assumption Human semantic labels, generated by LLMs and verified by human review, are correct ground truth for the intended meaning of each utterance.
    Both human and model evaluations are scored against these labels, so incorrect labels would invalidate the benchmark. Invoked in Sections 3.2.1 and 4.2.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DEBATE: A Dataset for Disentangling Textual Ambiguity in Mandarin Through Speech." pith.science (2026). https://pith.science/paper/PQDEYPYR

@misc{pith2026250607502,
  author       = {Pith},
  title        = {Pith review of: DEBATE: A Dataset for Disentangling Textual Ambiguity in Mandarin Through Speech},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PQDEYPYR}},
  note         = {Machine review of arXiv:2506.07502}
}
read the original abstract

Despite extensive research on textual and visual disambiguation, disambiguation through speech (DTS) remains underexplored. This is largely due to the lack of high-quality datasets that pair spoken sentences with richly ambiguous text. To address this gap, we present DEBATE, a unique public Chinese speech-text dataset designed to study how speech cues and patterns-pronunciation, pause, stress and intonation-can help resolve textual ambiguity and reveal a speaker's true intent. DEBATE contains 1,001 carefully selected ambiguous utterances, each recorded by 10 native speakers, capturing diverse linguistic ambiguities and their disambiguation through speech. We detail the data collection pipeline and provide rigorous quality analysis. Additionally, we benchmark three state-of-the-art large speech and audio-language models, illustrating clear and huge performance gaps between machine and human understanding of spoken intent. DEBATE represents the first effort of its kind and offers a foundation for building similar DTS datasets across languages and cultures. The dataset and associated code are available at: https://github.com/SmileHnu/DEBATE.

Figures

Figures reproduced from arXiv: 2506.07502 by the authors.

Figure 1
Figure 1. DEBATE is the first multimodal dataset designed for speech-based disambiguation in Mandarin Chinese, focusing on [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Pipeline for creating DEBATE—the first speech-text dataset designed for Chinese disambiguation through speech. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Illustration of the audio recording process. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Comparison of human and model performance on [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 42 canonical work pages

  1. [1]

    Oleg Akhtiamov, Dmitrii Ubskii, Evgeniia Feldina, Aleksei Pugachev, Alexey Kar- pov, and Wolfgang Minker. 2017. Are you addressing me? Multimodal addressee detection in human-human-computer conversations. In Proc. 19th International Conference on Speech and Computer (SPECOM) . Hatfield, UK, 152–161

    work page 2017

  2. [2]

    Keyu An, Qian Chen, Chong Deng, Zhihao Du, Changfeng Gao, Zhifu Gao, et al

    work page

  3. [3]

    Manjot Bedi, Shivani Kumar, Md Shad Akhtar, and Tanmoy Chakraborty. 2021. Multi-modal sarcasm detection and humor classification in code-mixed conver- sations. IEEE Transactions on Affective Computing 14, 2 (2021), 1363–1375

    work page 2021

  4. [4]

    Michele Bevilacqua, Tommaso Pasini, Alessandro Raganato, and Roberto Navigli

    work page

  5. [5]

    Balthasar Bickel and Johanna Nichols. 2007. Inflectional morphology. Language Typology and Syntactic Description 3, 2 (2007), 169–240. DEBATE: A Dataset for Disentangling Textual Ambiguity in Mandarin Through Speech Conference’17, July 2017, Washington, DC, USA

    work page 2007

  6. [6]

    Hui Bu, Jiayu Du, Xingyu Na, Bengu Wu, and Hao Zheng. 2017. AISHELL-1: An open-source mandarin speech corpus and a speech recognition baseline. In Proc. 20th Conference of the Oriental Chapter of the International Coordinating Commit- tee on Speech Databases and Speech I/O Systems and Assessment (O-COCOSDA) . Seoul, Korea, 1–5

    work page 2017

  7. [7]

    Chiara Bucaria. 2004. Lexical and syntactic ambiguity as a source of humor: The case of newspaper headlines. Humor 17, 3 (2004), 279–309

    work page 2004

  8. [8]

    Henry S Cheang and Marc D Pell. 2008. The sound of sarcasm. Speech Communi- cation 50, 5 (2008), 366–381

    work page 2008

Show all 46 references
  1. [9]

    Xinxiong Chen, Zhiyuan Liu, and Maosong Sun. 2014. A unified model for word sense representation and disambiguation. In Proc. Conference on Empirical Methods in Natural Language Processing (EMNLP) . Doha, Qatar, 1025–1035

    2014

  2. [10]

    Lukas Christ, Shahin Amiriparian, Alexander Kathan, Niklas Müller, Andreas König, and Björn W Schuller. 2025. Towards multimodal prediction of sponta- neous humor: A novel dataset and first results. IEEE Transactions on Affective Computing 16, 2 (2025), 844–860

    2025

  3. [11]

    Yunfei Chu, Jin Xu, Xiaohuan Zhou, Qian Yang, Shiliang Zhang, Zhijie Yan, et al. 2023. Qwen-audio: Advancing universal audio understanding via unified large-scale audio-language models. arXiv preprint arXiv:2311.07919 (2023)

    2023 arXiv

  4. [12]

    Google DeepMindg. 2024. Introducing Gemini 2.0: our new AI model for the agentic era. Retrieved May 30, 2025 from https://blog.google/technology/google- deepmind/google-gemini-ai-update-december-2024/

    2024

  5. [13]

    Soham Deshmukh, Benjamin Elizalde, Rita Singh, and Huaming Wang. 2023. Pengi: An audio language model for audio tasks. In Proc. Advances in Neural Information Processing Systems (NeurIPS), Vol. 36. New Orleans, LA, 18090–18108

    2023

  6. [14]

    Jiayu Du, Xingyu Na, Xuechen Liu, and Hui Bu. 2018. AISHELL-2: Transforming mandarin ASR research into industrial scale. arXiv preprint arXiv:1808.10583 (2018)

    2018 arXiv

  7. [15]

    Sreyan Ghosh, Sonal Kumar, Ashish Seth, Chandra Kiran Reddy Evuru, Utkarsh Tyagi, S Sakshi, et al. 2024. GAMA: A large audio-language model with advanced audio understanding and complex reasoning abilities. In Proc. Conference on Empirical Methods in Natural Language Processin...

    2024

  8. [16]

    Yuan Gong, Hongyin Luo, Alexander H Liu, Leonid Karlinsky, and James R Glass

  9. [17]

    Bairu Hou, Fanchao Qi, Yuan Zang, Xurui Zhang, Zhiyuan Liu, and Maosong Sun

  10. [18]

    Gaurav Kamath, Sebastian Schuster, Sowmya Vajjala, and Siva Reddy. 2024. Scope ambiguities in large language models. Transactions of the Association for Compu- tational Linguistics 12 (2024), 738–754

    2024

  11. [19]

    Listen, think, and understand. In Proc. 12th International Conference on Learning Representations (ICLR). Vienna, Austria, 30 pages

  12. [20]

    Charles N Li and Sandra A Thompson. 1989. Mandarin Chinese: A functional Reference Grammar. Univ of California Press

    1989

  13. [21]

    Alisa Liu, Zhaofeng Wu, Julian Michael, Alane Suhr, Peter West, Alexander Koller, et al. 2023. We’re Afraid Language Models Aren’t Modeling Ambiguity. In Proc. Conference on Empirical Methods in Natural Language Processing (EMNLP) . Singapore, 790–807

    2023

  14. [22]

    Esther Majambere. 2011. Clarity, precision and unambiguity: Aspects for effective legislative drafting. Commonwealth Law Bulletin 37, 3 (2011), 417–426

    2011

  15. [23]

    Bo Li, Huan Zhao, and Zixing Zhang. 2023. Diversifying emotional dialogue generation via selective adversarial training. Sensors 23, 13, Article 5904 (2023), 16 pages

    2023

  16. [24]

    Roberto Navigli, David Jurgens, and Daniele Vannella. 2013. SemEval-2013 Task 12: Multilingual Word Sense Disambiguation. In Proc. 7th International Workshop on Semantic Evaluation (SemEval) . Atlanta, GA, 222–231

    2013

  17. [25]

    Sunghyun Park, Han Suk Shim, Moitreya Chatterjee, Kenji Sagae, and Louis- Philippe Morency. 2016. Multimodal analysis and prediction of persuasiveness in online social multimedia. ACM Transactions on Interactive Intelligent Systems 6, 3, Article 25 (2016), 25 pages

    2016

  18. [26]

    Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, and Ilya Sutskever. 2023. Robust speech recognition via large-scale weak supervision. In Proc. International Conference on Machine Learning (ICML) . Honolulu, HI, 28492–28518

    2023

  19. [27]

    Andrea Moro and Roberto Navigli. 2015. SemEval-2015 Task 13: Multilingual All-Words Sense Disambiguation and Entity Linking. InProc. 9th International Workshop on Semantic Evaluation (SemEval) . Denver, CO, 288–297

    2015

  20. [28]

    Claudia Ross, Jing-heng Sheng Ma, Pei-Chia Chen, Baozhang He, and Meng Yeh

  21. [29]

    Björn Schuller, Stefan Steidl, Anton Batliner, Elika Bergelson, Jarek Krajewski, Christoph Janott, et al. 2017. The INTERSPEECH 2017 computational paralin- guistics challenge: Addressee, cold & snoring. In Proc. the Annual Conference of the International Speech Communication A...

    2017

  22. [30]

    Oswald Szemerényi and Oswald John Louis Szemerényi. 1999. Introduction to Indo-European Linguistics. OUP Oxford

    1999

  23. [31]

    Abhinav Rastogi, Xiaoxue Zang, Srinivas Sunkara, Raghav Gupta, and Pranav Khaitan. 2020. Towards scalable multi-domain conversational agents: The schema- guided dialogue dataset. In Proc. the AAAI Conference on Artificial Intelligence (AAAI). New York, NY, 8689–8696

    2020

  24. [32]

    Yue Wang, Qiliang Liang, Yaqi Yin, Hansi Wang, and Yang Liu. 2024. Disambiguate words like composing them: A morphology-informed approach to enhance Chi- nese word sense disambiguation. In Proc. 62nd Annual Meeting of the Association for Computational Linguistics (ACL). Bangko...

    2024

  25. [33]

    Routledge

    Modern Mandarin Chinese Grammar: A Practical Guide . Routledge

  26. [34]

    Jiaming Wu, Hongfei Lin, Liang Yang, and Bo Xu. 2021. Mumor: A multimodal dataset for humor detection in conversations. In Proc. 10th CCF International Conference on Natural Language Processing and Chinese Computing (NLPCC) . Qingdao, China, 619–627

    2021

  27. [35]

    Jin Xu, Zhifang Guo, Jinzheng He, Hangrui Hu, Ting He, Shuai Bai, et al. 2025. Qwen2. 5-omni technical report. arXiv preprint arXiv:2503.20215 (2025)

    2025 arXiv

  28. [36]

    Fiseha B Tesema, Jason Gu, Wei Song, Hong Wu, Shiqiang Zhu, Zheyuan Lin, et al. 2023. Addressee detection using facial and audio features in mixed human– human and human–robot settings: A deep learning framework. IEEE Systems, Man, and Cybernetics Magazine 9, 2 (2023), 25–38

    2023

  29. [37]

    Tan Yue, Xuzhao Shi, Rui Mao, Zonghai Hu, and Erik Cambria. 2024. SarcNet: A multilingual multimodal sarcasm detection dataset. In Proc. Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING). Turin, Italy, 14325–14335

    2024

  30. [38]

    Yue Wang, Hua Zheng, Yaqi Yin, Hansi Wang, Qiliang Liang, and Yang Liu. 2024. Morpheme Sense Disambiguation: A New Task Aiming for Understanding the Language at Character Level. In Proceedings of the Joint International Conference on Computational Linguistics, Language Resourc...

    2024

  31. [39]

    Zixing Zhang, Liyizhe Peng, Tao Pang, Jing Han, Huan Zhao, and Björn W Schuller. 2024. Refashioning emotion recognition modelling: The advent of generalised large models. IEEE Transactions on Computational Social Systems 11, 5 (2024), 6690–6704

    2024

  32. [40]

    Zixing Zhang, Weixiang Xu, Zhongren Dong, Kanglin Wang, Yimeng Wu, Jing Peng, et al . 2024. ParaLBench: A large-scale benchmark for computational paralinguistics over acoustic foundation models. IEEE Transactions on Affective Computing (2024). In press, 17 pages

    2024

  33. [41]

    Fukang Yan, Yue Zhang, and Zhenghua Li. 2023. Construction of a modern Chinese word sense dataset based on online dictionaries. In Proc. 22nd Chinese National Conference on Computational Linguistics . Harbin, China, 43–53

    2023

  34. [43]

    Qin Zhang, Sihan Cai, Jiaxu Zhao, Mykola Pechenizkiy, and Meng Fang. 2024. CHAmbi: A new benchmark on Chinese ambiguity challenges for large language models. In Proc. Findings of the Association for Computational Linguistics: EMNLP . Miami, FL, 14883–14898

    2024

  35. [46]

    Hua Zheng, Lei Li, Damai Dai, Deli Chen, Tianyu Liu, Xu Sun, and Yang Liu. 2021. Leveraging word-formation knowledge for Chinese word sense disambiguation. In Proc. Findings of the Association for Computational Linguistics: EMNLP . Punta Cana, Dominican Republic, 918–923

    2021

  36. [2020]

    Try to substitute: An unsupervised Chinese word sense disambiguation method based on Hownet. InProc. 28th International Conference on Computational Linguistics (COLING). Barcelona, Spain, 1752–1757

  37. [2021]

    Recent trends in word sense disambiguation: A survey. InProc. International Joint Conference on Artificial Intelligence (IJCAI) . Montreal, Canada, 4330–4338

  38. [2024]

    arXiv preprint arXiv:2407.04051 (2024)

    FunAudioLLM: Voice understanding and generation foundation models for natural interaction between humans and LLMs. arXiv preprint arXiv:2407.04051 (2024)

    2024 arXiv

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.