Pith. sign in

REVIEW 4 major objections 4 minor 39 references

Enhancing Persona Classification in Dialogue Systems: A Graph Neural Network Approach

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

Pith's one-line read This paper argues that a GraphSAGE head over a semantic-similarity persona graph improves multi-label persona classification, with the largest gains at 30–50% of the training data, and introduces a manually annotated dataset for this task.

desk verdict A new small persona dataset and a BertGCN-style GNN pipeline, but the low-data gains are likely inflated by transductive test-node access and test-set threshold tuning. read the letter →

arxiv 2412.13283 v1 pith:5RWF6ZDQ submitted 2024-12-17 cs.CL

classification cs.CL
keywords personaclassificationgraphneuralnetworkstextembeddingsmulti-labellow-resourcelearningSAGEnaturallanguageinferencedialoguesystems
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 argues that adding a graph neural network to a text-embedding classifier improves multi-label persona classification, and that the improvement is largest when labeled data is scarce. The authors build a weighted graph over persona statements—nodes are personas, edges connect the k=7 nearest neighbors by embedding cosine similarity, and edge weights come from NLI entailment scores—then train GraphSAGE jointly with fine-tuned bge-small embeddings. They also contribute a manually annotated dataset of about 3,000 personas drawn from the Multi-Session Chat corpus, labelled with five overlapping categories. The key result is that at 30% of the training data the combined model reaches F1 0.8325 versus 0.7127 for fine-tuned embeddings alone, while at 100% the margin narrows to 0.8872 versus 0.8742. The paper concludes that graph structure can substitute for labelled data in this task.

What carries the argument

The central object is a homogeneous, undirected, weighted graph whose nodes are persona statements. The construction fixes k=7 nearest neighbours for each node by cosine similarity of e5 embeddings, then reweights every resulting edge with the entailment score that a DeBERTa NLI classifier assigns to the pair; all edges, including weak ones, are kept. A GraphSAGE network consumes these node features and propagates information along the edges, and its output logits are merged with the fine-tuned encoder's logits through a weighted sum with λ=0.7, following the BertGCN design. The graph encodes the homophily assumption that semantically close personas are more likely to share labels, turning that assumption into a trainable inductive bias.

What would settle it

Rewire the graph edges randomly while keeping all node features and training the same FT-bge & GraphSAGE model; if the low-data F1 advantage over the linear-head baseline survives the rewiring, the reported gain is not caused by the graph structure. A complementary check is to measure edge homophily (the share of edges connecting same-label personas) across different k values and see whether the low-data gain tracks that share.

Watch

Extended reading notes

Core claim

The paper's central discovery is that relational structure among personas, expressed as a semantic-similarity graph, carries enough label information to substantially help a classifier when annotations are few. In the authors' experiments, the model that fine-tunes bge-small embeddings and trains a GraphSAGE head on the k-NN/NLI-weighted graph outperforms the same fine-tuned embeddings with a linear head at every training-set size, with the gap peaking at 30% of the data (F1 0.8325 vs 0.7127) and nearly disappearing at full data (0.8872 vs 0.8742). This is presented as evidence that the graph propagates label information from similar personas, which matters most exactly when direct supervision is limited.

Load-bearing premise

The pipeline assumes that personas that are semantically similar—by embedding cosine distance and NLI entailment—are more likely to carry the same labels, so the graph edges point the GNN toward useful neighbours; if that similarity-to-label agreement breaks, the graph adds noise rather than signal.

Editorial extensions

If this is right

  • At 30% of the training data, the GraphSAGE-augmented model reaches F1 0.8325, while fine-tuned embeddings alone reach 0.7127, so the graph closes more than half of the gap to the full-data result.
  • At 50% of the data, the combined model's F1 of 0.8826 already exceeds what fine-tuned embeddings achieve with 100% of the data (0.8742), implying the graph can substitute for a large share of labelled examples.
  • At 100% of the data, the two models are nearly tied (0.8872 vs 0.8742), so the graph's practical value is concentrated in low-resource settings.
  • The new manually annotated dataset of 2,889 training and 676 test personas across five overlapping classes gives the community a benchmark for persona classification that was previously missing.

Reading between the lines

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

  • A direct ablation—replacing the NLI edge weights with plain cosine similarities while keeping everything else fixed—would show whether the entailment weighting is the active ingredient or whether simple neighbourhood smoothing suffices.
  • The same recipe of k-NN graph construction plus a GNN head could transfer to other multi-label text classification tasks with scarce labels, such as dialogue act tagging or complaint categorisation, whenever a semantic-similarity graph can be built cheaply.
  • Because the dataset is built from a single dialogue corpus and pre-annotated by an LLM with a reported 20% error rate, the absolute F1 numbers are likely specific to this annotation setup; measuring how label noise propagates through the graph would clarify how much of the gain survives in noisier conditions.
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

4 major / 4 minor

Summary. The paper proposes a persona classification framework that combines text embeddings (bge-small) with a GraphSAGE model over a homogeneous graph whose nodes are persona statements and whose edges are k-NN similarities weighted by an NLI entailment model. The authors construct a manually validated dataset of about 3,565 personas from the MSC corpus, using LLM-assisted annotation with human correction, and evaluate multi-label F1 at 1%, 30%, 50%, 70%, and 100% of the training data. They report that fine-tuned embeddings plus GraphSAGE outperform fine-tuned embeddings alone, especially at 30--50% data, and conclude that GNNs are valuable under data scarcity.

Significance. If the central claim held, the contribution would be useful: a new manually annotated persona-classification dataset from MSC and a demonstration that relational structure among persona statements improves low-resource multi-label classification. The paper has the merit of attacking a real data bottleneck and of combining a k-NN graph with NLI-based edge weights rather than relying on raw lexical statistics. However, the evaluation as reported does not currently establish the claimed advantage: the GNN evaluation is transductive while the baselines are not, the per-model probability thresholds are tuned on the test set, and the low-data comparisons lack significance testing. The dataset itself is a contribution, but its utility depends on release and on documentation of annotation reliability.

major comments (4)
  1. [§4.1 and §4.2.4] The reported low-data gains may be an artifact of transductive test-node inclusion. The graph in Section 4.1 is constructed from the full persona set and Section 4.2.4 states that the GNN 'needs to be trained on all data samples' following BertGCN; no inductive evaluation is described. Test personas can therefore contribute edges and receive propagated label information during training, whereas the FT-bge & Linear baseline never sees test instances. This asymmetric setup can explain much of the 30% and 50% gaps in Table 4 (0.8325 vs 0.7127 and 0.8826 vs 0.8330). Please add an inductive control with test nodes removed from the graph during training or, at minimum, compare against a transductive label-propagation baseline that is allowed the same graph access.
  2. [§4.2.5] The per-model decision threshold is selected on the test set: 'For each model, we identify the optimal threshold for the probabilities, which corresponds to the highest F1 score.' This optimizes the reported metric on the evaluation set for each method separately; because model confidence distributions differ, threshold tuning is an uncontrolled advantage and can systematically inflate the reported F1, particularly for the 1% runs where FT-bge & Linear reaches recall 1.0 with F1 0.474. Thresholds should be fixed on a validation split, or a threshold-free metric such as average precision should be reported.
  3. [Table 4 and §5] The claim that GNN integration 'significantly improves classification performance, especially with limited data' is not backed by significance tests, and the 1% results do not support it. At 1%, Pre-bge & GraphSAGE (0.5054±0.0086) overlaps with Pre-bge & Linear (0.4582±0.1610), and FT-bge & GraphSAGE (0.4770±0.0100) overlaps with FT-bge & Linear (0.4740±0.0000). At 30% and 50% the means differ, but the paper reports no paired tests over the 10 runs. Please report paired bootstrap or signed-rank tests for the FT-bge & GraphSAGE versus FT-bge & Linear comparison at each data fraction, and be careful not to claim significance without such tests.
  4. [§3.1 and §4.1] The homophily assumption stated in Section 3.1 — 'Since semantically close personas are more likely to have the same classes' — is load-bearing but not validated. The graph edges are defined by e5 cosine similarity and DeBERTa NLI entailment scores, not by labels, so the usefulness of the graph for label propagation is an empirical premise. Add an analysis of edge-label homophily (e.g., the fraction of k-NN neighbors sharing at least one label) and a control with random edges or embedding-only similarity to confirm that the reported gains come from the graph structure rather than from the additional GNN capacity.
minor comments (4)
  1. [§4.2.5] The sentence 'we select the best model based on the highest F1 score' is ambiguous about whether model selection is done on a validation set or on the test set; please specify the validation procedure explicitly.
  2. [Table 3] The per-label counts sum to more than the 'Overall' row because of multi-label annotation; state the number of unique personas and the average number of labels per persona to make the dataset statistics interpretable.
  3. [Table 4 and Appendix B] BOW results are described as baselines in §4.2.1 but omitted from Table 4; either include them in the main table or refer explicitly to Appendix B. Also, the 100% row 'Pre-bge GraphSAGE' is missing the ampersand used elsewhere.
  4. [§3.1] The statement that the taxonomy follows PeaCoK is imprecise because PeaCoK defines relation types between personas, whereas this paper classifies nodes; clarifying this distinction would help the reader assess the novelty of the label scheme.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the GNN gains are supported by an empirical comparison whose graph inputs are label-independent.

full rationale

The paper's central claim is an empirical one (GNN integration improves persona classification, especially with limited data), and the supporting experiments do not reduce to the paper's own inputs by construction. The graph is built from e5 text embeddings and DeBERTa NLI entailment scores, not from the label distribution, so the GNN is not trivially receiving the answer. The final logits in Eq. (4) combine encoder and GNN outputs through a fixed weighted sum, and neither k=7 nor lambda=0.7 is fitted to reproduce the reported F1 numbers; lambda is imported from BertGCN as an external design choice. The only imported conceptual component is the PeaCoK label taxonomy, which is prior external work (not self-citation) and is reused as both annotation scheme and evaluation target; that is a benchmark convention, not a circular derivation. The transductive training protocol described in Section 4.2.4, where the graph includes all personas and the GNN is trained on all data samples, is a legitimate experimental-design concern that could inflate the apparent low-data advantage, but test labels are never used as inputs, so it is a leakage/validity risk rather than circularity. No equation in the paper equates a prediction to a fitted parameter or to a self-cited result, and the Limitations section does not concede any circular dependency. Under the required standard of quoting a specific reduction, no circular step is present.

Assumptions & free parameters 3 free parameters · 2 assumptions · 0 invented entities

The central empirical claim depends mainly on hyperparameters (k, lambda, and a per-model test-tuned threshold) and on two domain assumptions about label semantics and annotation quality. No invented entities or new theoretical postulates are introduced.

free parameters (3)
  • k (number of graph neighbors) = 7
    Chosen as optimal in the experiments (Section 4.1); alters graph density and the GNN's neighborhood aggregation.
  • lambda (GNN logit weight) = 0.7
    Set to 0.7 following BertGCN (Section 4.2.4); controls the blend of encoder and GNN outputs.
  • probability threshold per model = not reported (optimized)
    For each model the threshold is tuned on the test set to maximize F1 (Section 4.2.5), which can inflate reported F1 scores.
assumptions (2)
  • domain assumption Semantically similar personas tend to share the same class labels
    The graph construction and GNN rely on this homophily assumption; stated in Section 3.1: 'Since semantically close personas are more likely to have the same classes...' If false, the graph adds noise.
  • domain assumption The human-validated LLM annotations are accurate ground truth
    The dataset labels come from LLM output plus manual validation; LLM errors were found in 20% of cases before validation (Section 6), so residual label noise is possible.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhancing Persona Classification in Dialogue Systems: A Graph Neural Network Approach." pith.science (2026). https://pith.science/paper/5RWF6ZDQ

@misc{pith2026241213283,
  author       = {Pith},
  title        = {Pith review of: Enhancing Persona Classification in Dialogue Systems: A Graph Neural Network Approach},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5RWF6ZDQ}},
  note         = {Machine review of arXiv:2412.13283}
}
read the original abstract

In recent years, Large Language Models (LLMs) gain considerable attention for their potential to enhance personalized experiences in virtual assistants and chatbots. A key area of interest is the integration of personas into LLMs to improve dialogue naturalness and user engagement. This study addresses the challenge of persona classification, a crucial component in dialogue understanding, by proposing a framework that combines text embeddings with Graph Neural Networks (GNNs) for effective persona classification. Given the absence of dedicated persona classification datasets, we create a manually annotated dataset to facilitate model training and evaluation. Our method involves extracting semantic features from persona statements using text embeddings and constructing a graph where nodes represent personas and edges capture their similarities. The GNN component uses this graph structure to propagate relevant information, thereby improving classification performance. Experimental results show that our approach, in particular the integration of GNNs, significantly improves classification performance, especially with limited data. Our contributions include the development of a persona classification framework and the creation of a dataset.

Figures

Figures reproduced from arXiv: 2412.13283 by the authors.

Figure 1
Figure 1. An example of a subgraph. Here the blue color is the label "Experiences", the red color is the multilabel "Experiences" and "Characteristics". The model is designed to classify text by lever￾aging both a pre-trained text encoder and a GNN. The encoder captures the semantic content of the text, while the GNN incorporates graph-based re￾lational information. The output logits from both components are combined using a … view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 13 canonical work pages

  1. [1]

    Aguda, Suchetha Siddagangappa, Elena Kochkina, Simerjot Kaur, Dongsheng Wang, and Charese Smiley

    Toyin D. Aguda, Suchetha Siddagangappa, Elena Kochkina, Simerjot Kaur, Dongsheng Wang, and Charese Smiley. 2024. https://aclanthology.org/2024.lrec-main.885 Large language models as financial data annotators: A study on effectiveness and efficiency . In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources...

  2. [2]

    Jiale Cheng, Sahand Sabour, Hao Sun, Zhuang Chen, and Minlie Huang. 2023. https://doi.org/10.18653/v1/2023.findings-acl.34 PAL : Persona-augmented emotional support conversation generation . In Findings of the Association for Computational Linguistics: ACL 2023, pages 535--554, Toronto, Canada. Association for Computational Linguistics

  3. [3]

    Bijoyan Das and Sarit Chakraborty. 2018. https://arxiv.org/abs/1806.06407 An improved text sentiment classification model using tf-idf and next word negation . Preprint, arXiv:1806.06407

  4. [4]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. https://doi.org/10.18653/v1/N19-1423 BERT : Pre-training of deep bidirectional transformers for language understanding . In Proceedings of the 2019 Conference of the North A merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long a...

  5. [5]

    Lukas Galke and Ansgar Scherp. 2022. https://doi.org/10.18653/v1/2022.acl-long.279 Bag-of-words vs. graph vs. sequence in text classification: Questioning the necessity of text-graphs and the surprising strength of a wide MLP . In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 4038--4...

  6. [6]

    Silin Gao, Beatriz Borges, Soyoung Oh, Deniz Bayazit, Saya Kanno, Hiromi Wakaki, Yuki Mitsufuji, and Antoine Bosselut. 2023. https://doi.org/10.18653/v1/2023.acl-long.362 P ea C o K : Persona commonsense knowledge for consistent and engaging narratives . In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: ...

  7. [7]

    Garrido-Merchan, Roberto Gozalo-Brizuela, and Santiago Gonzalez-Carvajal

    Eduardo C. Garrido-Merchan, Roberto Gozalo-Brizuela, and Santiago Gonzalez-Carvajal. 2023. https://doi.org/10.47852/bonviewjcce3202838 Comparing bert against traditional machine learning models in text classification . Journal of Computational and Cognitive Engineering, 2(4):352–356

  8. [8]

    Hamilton, Rex Ying, and Jure Leskovec

    William L. Hamilton, Rex Ying, and Jure Leskovec. 2017. https://arxiv.org/abs/1706.02216 Inductive representation learning on large graphs . CoRR, abs/1706.02216

Show all 39 references
  1. [9]

    Jiabang He, Liu Jia, Lei Wang, Xiyao Li, and Xing Xu. 2023. https://arxiv.org/abs/2308.08204 Mocosa: Momentum contrast for knowledge graph completion with structure-augmented pre-trained language models . Preprint, arXiv:2308.08204

  2. [10]

    Weihua Hu, Matthias Fey, Marinka Zitnik, Yuxiao Dong, Hongyu Ren, Bowen Liu, Michele Catasta, and Jure Leskovec. 2020. https://arxiv.org/abs/2005.00687 Open graph benchmark: Datasets for machine learning on graphs . CoRR, abs/2005.00687

  3. [11]

    Qian Huang, Horace He, Abhay Singh, Ser-Nam Lim, and Austin Benson. 2021. https://openreview.net/forum?id=8E1-f3VhX1o Combining label propagation and simple models out-performs graph neural networks . In International Conference on Learning Representations

  4. [12]

    Shenyang Huang, Farimah Poursafaei, Jacob Danovitch, Matthias Fey, Weihua Hu, Emanuele Rossi, Jure Leskovec, Michael Bronstein, Guillaume Rabusseau, and Reihaneh Rabbany. 2023. https://arxiv.org/abs/2307.01026 Temporal graph benchmark for machine learning on temporal graphs . ...

  5. [13]

    Mohammad Rasool Izadi, Yihao Fang, Robert Stevenson, and Lizhen Lin. 2020. https://doi.org/10.1109/BigData50022.2020.9378063 Optimization of graph neural networks with natural gradient descent . In 2020 IEEE International Conference on Big Data (Big Data), pages 171--179

  6. [14]

    Dongshi Ju, Shi Feng, Pengcheng Lv, Daling Wang, and Yifei Zhang. 2022. https://aclanthology.org/2022.coling-1.23 Learning to improve persona consistency in multi-party dialogue generation via text knowledge enhancement . In Proceedings of the 29th International Conference on ...

  7. [15]

    Tomohito Kasahara, Daisuke Kawahara, Nguyen Tung, Shengzhe Li, Kenta Shinzato, and Toshinori Sato. 2022. https://doi.org/10.18653/v1/2022.naacl-srw.13 Building a personalized dialogue system with prompt-tuning . In Proceedings of the 2022 Conference of the North American Chapt...

  8. [16]

    Yova Kementchedjhieva and Ilias Chalkidis. 2023. https://arxiv.org/abs/2305.05627 An exploration of encoder-decoder approaches to multi-label classification for legal and biomedical text . Preprint, arXiv:2305.05627

  9. [17]

    Kipf and Max Welling

    Thomas N. Kipf and Max Welling. 2016. https://arxiv.org/abs/1609.02907 Semi-supervised classification with graph convolutional networks . CoRR, abs/1609.02907

  10. [18]

    Shad Akhtar, and Tanmoy Chakraborty

    Shivani Kumar, Rishabh Gupta, Md. Shad Akhtar, and Tanmoy Chakraborty. 2024. https://aclanthology.org/2024.lrec-main.37 Adding SPICE to life: Speaker profiling in multiparty conversations . In Proceedings of the 2024 Joint International Conference on Computational Linguistics,...

  11. [19]

    Deuksin Kwon, Sunwoo Lee, Ki Hyun Kim, Seojin Lee, Taeyoon Kim, and Eric Davis. 2023. https://doi.org/10.18653/v1/2023.acl-industry.68 What, when, and how to ground: Designing user persona-aware conversational agents for engaging dialogue . In Proceedings of the 61st Annual Me...

  12. [20]

    Haotian Li, Bin Yu, Yuliang Wei, Kai Wang, Richard Yi Da Xu, and Bailing Wang. 2024. https://arxiv.org/abs/2309.14770 Kermit: Knowledge graph completion of enhanced relation modeling with inverse transformation . Preprint, arXiv:2309.14770

  13. [21]

    Sha Li, Mahdi Namazifar, Di Jin, Mohit Bansal, Heng Ji, Yang Liu, and Dilek Hakkani-Tur. 2022. https://doi.org/10.18653/v1/2022.naacl-main.202 Enhancing knowledge selection for grounded dialogues via document semantic graphs . In Proceedings of the 2022 Conference of the North...

  14. [22]

    Yu-Chen Lin, Si-An Chen, Jie-Jyun Liu, and Chih-Jen Lin. 2023. https://doi.org/10.18653/v1/2023.acl-short.160 Linear classifier: An often-forgotten baseline for text classification . In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Vo...

  15. [23]

    Yuxiao Lin, Yuxian Meng, Xiaofei Sun, Qinghong Han, Kun Kuang, Jiwei Li, and Fei Wu. 2021. https://doi.org/10.18653/v1/2021.findings-acl.126 B ert GCN : Transductive text classification by combining GNN and BERT . In Findings of the Association for Computational Linguistics: A...

  16. [24]

    Amirhossein Nouranizadeh, Mohammadjavad Matinkia, Mohammad Rahmati, and Reza Safabakhsh. 2021. https://arxiv.org/abs/2107.01410 Maximum entropy weighted independent set pooling for graph neural networks . CoRR, abs/2107.01410

  17. [25]

    Qader, Musa M

    Wisam A. Qader, Musa M. Ameen, and Bilal I. Ahmed. 2019. https://doi.org/10.1109/IEC47844.2019.8950616 An overview of bag of words;importance, implementation, applications, and challenges . In 2019 International Engineering Conference (IEC), pages 200--204

  18. [26]

    Hamidreza Rouzegar and Masoud Makrehchi. 2024. https://aclanthology.org/2024.law-1.10 Enhancing text classification through LLM -driven active learning and human annotation . In Proceedings of The 18th Linguistic Annotation Workshop (LAW-XVIII), pages 98--111, St. Julians, Mal...

  19. [27]

    Chi Sun, Xipeng Qiu, Yige Xu, and Xuanjing Huang. 2020. https://arxiv.org/abs/1905.05583 How to fine-tune bert for text classification? Preprint, arXiv:1905.05583

  20. [28]

    Zhen Tan, Dawei Li, Song Wang, Alimohammad Beigi, Bohan Jiang, Amrita Bhattacharjee, Mansooreh Karami, Jundong Li, Lu Cheng, and Huan Liu. 2024. https://arxiv.org/abs/2402.13446 Large language models for data annotation: A survey . Preprint, arXiv:2402.13446

  21. [29]

    Yihong Tang, Bo Wang, Miao Fang, Dongming Zhao, Kun Huang, Ruifang He, and Yuexian Hou. 2023. https://doi.org/10.18653/v1/2023.acl-long.299 Enhancing personalized dialogue generation with contrastive latent variables: Combining sparse and dense persona . In Proceedings of the ...

  22. [30]

    Petar Veličković, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Liò, and Yoshua Bengio. 2018. https://arxiv.org/abs/1710.10903 Graph attention networks . Preprint, arXiv:1710.10903

  23. [31]

    Liang Wang, Wei Zhao, Zhuoyu Wei, and Jingming Liu. 2022. https://doi.org/10.18653/v1/2022.acl-long.295 S im KGC : Simple contrastive knowledge graph completion with pre-trained language models . In Proceedings of the 60th Annual Meeting of the Association for Computational Li...

  24. [32]

    Jing Xu, Arthur Szlam, and Jason Weston. 2022. https://doi.org/10.18653/v1/2022.acl-long.356 Beyond goldfish memory: Long-term open-domain conversation . In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 5...

  25. [33]

    Xinchao Xu, Zeyang Lei, Wenquan Wu, Zheng-Yu Niu, Hua Wu, and Haifeng Wang. 2023. https://doi.org/10.18653/v1/2023.findings-acl.90 Towards zero-shot persona dialogue generation with in-context learning . In Findings of the Association for Computational Linguistics: ACL 2023, p...

  26. [34]

    Liang Yao, Chengsheng Mao, and Yuan Luo. 2018. https://arxiv.org/abs/1809.05679 Graph convolutional networks for text classification . Preprint, arXiv:1809.05679

  27. [35]

    Saizheng Zhang, Emily Dinan, Jack Urbanek, Arthur Szlam, Douwe Kiela, and Jason Weston. 2018. https://doi.org/10.18653/v1/P18-1205 Personalizing dialogue agents: I have a dog, do you have pets too? In Proceedings of the 56th Annual Meeting of the Association for Computational ...

  28. [36]

    Zhen Zhang, Jiajun Bu, Martin Ester, Jianfeng Zhang, Zhao Li, Chengwei Yao, Huifen Dai, Zhi Yu, and Can Wang. 2023. https://doi.org/10.1109/TKDE.2021.3090664 Hierarchical multi-view graph pooling with structure learning . IEEE Transactions on Knowledge and Data Engineering, 35...

  29. [37]

    Luyao Zhu, Wei Li, Rui Mao, Vlad Pandelea, and Erik Cambria. 2023. https://doi.org/10.18653/v1/2023.acl-long.544 PAED : Zero-shot persona attribute extraction in dialogues . In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: L...

  30. [38]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list...

  31. [39]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

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