Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

Beyond Whole Dialogue Modeling: Contextual Disentanglement for Conversational Recommendation

T0 review · 4 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read Separating focus from background information in dialogue improves conversational recommendations.

desk verdict A plausible new CRS model that separates dialogue context into focus and background, with consistent empirical wins, but the disentanglement mechanism needs a clearer definition and a less circular loss. read the letter →

arxiv 2504.17427 v1 pith:T6UILJ75 submitted 2025-04-24 cs.IR

Guojia An , Jie Zou , Jiwei Wei , Chaoning Zhang , Fuming Sun , Yang Yang This is my paper · ORCID
classification cs.IR
keywords conversationalrecommendationcontextualdisentanglementfocusandbackgroundinformationcontrastivelearningcounterfactualinferenceadaptivepromptknowledgegraph
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

Conversational recommender systems usually encode the entire dialogue history as one block, mixing the things a user mentions with the reasons and context around those mentions. This paper argues that this mixing is a bottleneck and that the two kinds of information—focus (entity-related) and background (occasion, companion, tone)—should be disentangled first and then recombined according to the situation. The proposed model, DisenCRS, separates focus and background representations with a contrastive loss and a counterfactual inference loss, then uses a learnable prompt selector to pick the best weighted fusion for each dialogue. On the ReDial and INSPIRED datasets the paper reports consistent gains over whole-dialogue baselines on recommendation metrics (Recall, NDCG, MRR) and generation metrics (BLEU, ROUGE, Distinct). The claim that matters is that contextual disentanglement plus adaptive prompt selection is a broadly useful design for conversational recommendation, not just a trick for these two datasets.

What carries the argument

The dual contextual disentanglement module is the load-bearing mechanism: a contrastive triplet loss that pulls focus information h_f toward entity-derived proxy h_p and pushes it away from background h_b, paired with a counterfactual loss that uses similarity to the ground-truth item to decide which factor dominated the user's choice and reinforces that dominance. Around this sits the adaptive prompt learning module, which builds a prompt pool of weighted fusions of h_f, h_b, and the dialogue context, then uses a learnable multilayer-perceptron selector to pick the best prompt for the current dialogue. Frozen DialoGPT consumes the selected prompt for both downstream tasks.

What would settle it

Have human annotators label which spans of a held-out set of ReDial and INSPIRED dialogues are focus (entity-related) and which are background; if DisenCRS's h_f and h_b do not align with those labels better than a random split of the dialogue representation, the disentanglement is not achieving the separation the paper claims.

Watch

Extended reading notes

Core claim

The central discovery is that dialogue context in conversational recommendation is not homogeneous: it mixes focus information (entities and entity-related preferences) with background information (situational cues such as who the user is with or the tone of the request), and modeling both as one vector lowers recommendation accuracy. DisenCRS extracts a focus representation h_f and a background representation h_b from the dialogue's RoBERTa [CLS] embedding, then supervises their separation without manual labels. Contrastive disentanglement (Eq. 8) uses entity-derived semantic representation h_p as a proxy and a triplet-style loss to make focus closer to the proxy than to background. Counterfactual inference disentanglement (Eq. 9) compares h_f and h_b to the ground-truth item, assigns a pseudo-label for which factor dominated the choice, and uses that signal to further separate the two. The separated representations are then fused into a prompt pool and an MLP selector chooses the best weighted prompt for each dialogue, with the selected prompt fed to a frozen DialoGPT for both recommendation and response generation. Ablations in the paper show that removing either disentanglement mechanism or replacing adaptive selection with fixed weights hurts performance.

Load-bearing premise

The load-bearing premise is that entity mentions are a valid proxy for focus information and that similarity to the ground-truth item tells which factor dominated the user's choice, so the separation is guided by correlation-based pseudo-labels generated from the same representations being trained, not by an independent definition of focus and background.

Editorial extensions

If this is right

  • Other conversational recommender models could adopt the same separation step and expect gains, since the paper shows the disentanglement module improves three existing whole-dialogue models.
  • Background information such as occasion, companions, or constraints should be treated as a signal to weigh, not as noise to discard.
  • Adaptive prompt selection beats any fixed weighting of focus and background, so the right balance varies from dialogue to dialogue.
  • Improvements appear on both item recommendation and response generation, meaning the disentanglement also changes what the system says, not just what it recommends.

Reading between the lines

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

  • The entity-mention proxy could be replaced by LLM-generated annotations of focus and background; if gains persist, the method's value is in the separation itself rather than in the specific proxy.
  • Because the counterfactual loss leans on the ground-truth item to decide which factor dominated, the method may inherit popularity bias from the datasets; testing on a popularity-controlled candidate set would show whether disentanglement helps hard cases.
  • The same focus/background split may transfer to other interactive settings such as travel planning or e-commerce, where the user's occasion and constraints are background and product attributes are focus.
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 / 5 minor

Summary. The paper proposes DisenCRS, a conversational recommender system that separates dialogue context into focus information (entity-related) and background information (non-entity), using a dual disentanglement module composed of contrastive disentanglement and counterfactual inference disentanglement. The disentangled representations are then fed into an adaptive prompt learning module that selects from a prompt pool with pre-defined weight combinations, and the selected prompt is used with frozen DialoGPT for both item recommendation and response generation. Experiments on ReDial and INSPIRED report consistent improvements over baselines on Recall, NDCG, MRR, BLEU, ROUGE, and Distinct metrics, plus a small human evaluation.

Significance. If the central claim holds, the paper offers a plausible new angle for CRS: rather than treating the whole dialogue as a single representation, explicitly separating focus and background information and adaptively weighting them could improve both recommendation and generation. The empirical gains in Tables 2 and 3 are consistent across many metrics and two datasets, which is a genuine strength. The paper also provides ablation studies for the disentanglement module and the adaptive prompt selection, and a hyperparameter study for the number of prompts and the loss weight. However, the conceptual contribution depends entirely on whether the disentanglement mechanism actually separates interpretable factors; as written, the derivation of h_f and h_b is underspecified and the losses may be fitting to the same prediction target, so the significance is conditional on a technical clarification that the current manuscript does not provide.

major comments (4)
  1. [Section 3.3, Eqs. (6)-(9)] The core disentanglement is underspecified: h_f and h_b are introduced as being 'disentangled from the dialogue semantic representation h_cls' but no extraction equation is given. There is no projection, gate, mask, or nonlinear transformation that defines h_f and h_b as functions of h_cls. Without such an equation, the reader cannot verify that the two vectors are different views rather than two identical copies of h_cls, and the subsequent contrastive and counterfactual losses are the only optimization signals that can differentiate them.
  2. [Eq. (9) and Eq. (21)] The counterfactual loss is circular with respect to the recommendation target: Eq. (9) builds pseudo-labels D_f by comparing h_f and h_b with the ground-truth item h_t, then optimizes the similarity between the 'dominant' factor and h_t. The same h_t is the supervision signal in the recommendation loss Eq. (21). This means the disentanglement signal is fitted to the prediction target itself, so the module may simply learn two projections of the same predictive signal rather than focus versus background. The paper should provide an independent definition of focus/background or an out-of-sample test (e.g., whether the learned factors predict held-out entity mentions versus non-entity context).
  3. [Section 4.3, Figure 4] The ablation study for the disentanglement module is reported only as a figure without numerical values, error bars, or significance tests. Given that the central claim is that disentanglement itself provides the gains, the reader needs to see the magnitude of the drops for '- w/o CD', '- w/o CID', and '- w/o Dual', and ideally confidence intervals, to assess whether the improvements are meaningful or within noise.
  4. [Section 4.4, Table 5] The comparison 'DisenCRS-fw' (fixed weights) is said to use a grid search for the best manual weights, but the selected weights are not reported. Without knowing what fixed weights were chosen, the reader cannot judge whether the adaptive selector is genuinely better than a well-tuned static fusion, or whether the grid search was too coarse. Reporting the grid range and the best fixed weights would make the comparison reproducible.
minor comments (5)
  1. [Eq. (8) and surrounding text] Eq. (8) is written as a max over a similarity difference, but the text describes a triplet loss based on Euclidean distance; the notation is inconsistent. Also, the expression sim(x,y) = x·y / (||x|| ||y||) has a typo in the denominator (missing the second norm).
  2. [Section 3.3.2, around Eq. (9)] The phrase 'generating pseudo labels for each conversation leader' is unclear; likely 'conversation' or 'conversation turn' is meant. Also, the definition of D_f is introduced in the equation but never formally defined in the text; it should be defined explicitly as the set of conversations where focus dominates.
  3. [Section 4.1.3, baseline list] The baseline list mentions BERT, GPT-2, and BART, but Table 2 only reports results for BERT and not for GPT-2 or BART on the recommendation task; the paper should clarify which baselines are used for which task, or include the missing columns.
  4. [Section 4.5.1, Figure 5] Figure 5 appears to be a line plot without axis labels or a legend in the extracted text; the y-axis metric and the curve identifiers are missing. The hyperparameter study would be much clearer if the figure included labeled axes and curves for the reported metrics.
  5. [Section 4.5.2, Figure 6] Similar to Figure 5, Figure 6 lacks axis labels and a legend in the extracted text; the reader cannot tell which metric is plotted against lambda. Please add clear labels.

Circularity Check

1 steps flagged · score 3.0 of 10

The counterfactual disentanglement loss is self-referential: the pseudo-label and the objective both use the same similarity between h_f/h_b and the ground-truth item; otherwise the central empirical claim is benchmarked externally.

  1. self definitional [Section 3.3.2, Eq. (9), 'Counterfactual Inference Disentanglement']
    "we compute the similarity between the focal information and background information with the ground truth item respectively, thus generating pseudo labels for each conversation leader. Then Based on the idea of counterfactual inference, the similarity between the dominant factor and the target item is expected to be significantly greater than the similarity between the other factor and the target item. L_ci = { - sim(h_f,h_t)/(sim(h_f,h_t)+sim(h_b,h_t)), h_t in D_f; - sim(h_b,h_t)/(sim(h_b,h_t)+sim(h_f,h_t)), h_t not in D_f }"

    The branch condition D_f is not fixed by any external definition; it is generated by comparing sim(h_f,h_t) with sim(h_b,h_t), and the loss then minimizes the negative normalized similarity of whichever factor is currently closer to h_t. Thus L_ci has no independent supervisory content: it reinforces the initial winner of the same comparison that defines the pseudo-label. Moreover h_t is the same ground-truth item optimized in Eq. (21), so the 'disentanglement' signal is directly coupled to the prediction target rather than to an independent notion of focus versus background. The counterfactual intervention (removing one factor) is never actually computed; the loss reduces to ranking by similarity to the target item.

full rationale

Most of the paper is an empirical systems paper, so the main circularity axes do not apply: the recommendation and generation results are evaluated against external baselines on public datasets (ReDial and INSPIRED), and the prompt-learning components have independent content. The one definitional loop is in the 'counterfactual' disentanglement loss: the pseudo-label D_f is produced by the same similarity comparison between h_f/h_b and the ground-truth item h_t that the loss then maximizes, so L_ci is a self-reinforcing ranking objective rather than a counterfactual intervention. Because h_t also appears in L_rec, the disentanglement signal is coupled to the prediction target; however, this does not force the headline results, since the final comparisons are against independently trained baselines and the ablation in Figure 4 provides some evidence that the module contributes. The lack of explicit equations defining h_f and h_b, and the absence of code and error bars, are reproducibility/correctness concerns rather than circularity. The score of 3 reflects the one self-referential step while recognizing that the central empirical claim is externally grounded.

Assumptions & free parameters 4 free parameters · 4 assumptions · 2 invented entities

The central claim rests on two invented latent factors (focus and background) that are never directly validated. Their supervision comes from entity proxies and from similarity to the ground-truth item, so the ledger is dominated by unverified constructs and tuned hyperparameters rather than independent evidence.

free parameters (4)
  • triplet margin m = not reported
    Controls the minimum similarity gap in L_cd; no value or tuning range is given.
  • disentanglement loss weight lambda = not reported (tuned in range 1.0 to 7.0)
    Scales L_cd + L_ci in Eq. (20); Fig. 6 shows it materially changes performance.
  • number of prompts eta = not reported (best near 10 by Fig. 5)
    Sets the granularity of the prompt pool; tuned as a hyperparameter.
  • prompt pool weights W_i_f, W_i_b = predefined grid, sum to 1
    Manually chosen weight pairs in Eq. (13); the selector picks an index among them.
assumptions (4)
  • domain assumption Dialogue context can be decomposed into focus and background information
    Stated in the introduction and used throughout; no formal definition or annotations.
  • ad hoc to paper Entity mentions are a valid proxy for focus information
    Contrastive loss in Eq. (8) pulls h_f toward h_p, the semantic entity representation, assuming entity-related content equals user focus.
  • ad hoc to paper Similarity between a latent factor and the ground-truth item indicates causal dominance
    Eq. (9) assigns pseudo-labels D_f based on sim(h_f,h_t) vs sim(h_b,h_t); this equates correlation with influence.
  • standard math Pretrained models RoBERTa and DialoGPT provide adequate semantic encodings
    Used as frozen feature extractors, standard practice in the cited baselines.
invented entities (2)
  • Focus information latent vector h_f
    purpose: Carries entity-related user intent disentangled from the dialogue
    No annotation or behavioral test verifies that h_f isolates focus; only downstream recommendation performance is reported.
  • Background information latent vector h_b
    purpose: Carries non-entity context such as social situation
    Same as h_f; no independent measurement.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Beyond Whole Dialogue Modeling: Contextual Disentanglement for Conversational Recommendation." pith.science (2026). https://pith.science/paper/T6UILJ75

@misc{pith2026250417427,
  author       = {Pith},
  title        = {Pith review of: Beyond Whole Dialogue Modeling: Contextual Disentanglement for Conversational Recommendation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/T6UILJ75}},
  note         = {Machine review of arXiv:2504.17427}
}
read the original abstract

Conversational recommender systems aim to provide personalized recommendations by analyzing and utilizing contextual information related to dialogue. However, existing methods typically model the dialogue context as a whole, neglecting the inherent complexity and entanglement within the dialogue. Specifically, a dialogue comprises both focus information and background information, which mutually influence each other. Current methods tend to model these two types of information mixedly, leading to misinterpretation of users' actual needs, thereby lowering the accuracy of recommendations. To address this issue, this paper proposes a novel model to introduce contextual disentanglement for improving conversational recommender systems, named DisenCRS. The proposed model DisenCRS employs a dual disentanglement framework, including self-supervised contrastive disentanglement and counterfactual inference disentanglement, to effectively distinguish focus information and background information from the dialogue context under unsupervised conditions. Moreover, we design an adaptive prompt learning module to automatically select the most suitable prompt based on the specific dialogue context, fully leveraging the power of large language models. Experimental results on two widely used public datasets demonstrate that DisenCRS significantly outperforms existing conversational recommendation models, achieving superior performance on both item recommendation and response generation tasks.

Figures

Figures reproduced from arXiv: 2504.17427 by the authors.

Figure 2
Figure 2. Performance comparison of the original method [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. The overall architecture of DisenCRS, which consists of four main components: (1) the representation learning module, [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Ablation comparison of the dual contextual disen [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Comparison of the number of prompts 𝜂 in the prompt pool. 1.0 2.0 3.0 4.0 5.0 6.0 7.0 0.06 0.07 0.08 0.09 R@1 1.0 2.0 3.0 4.0 5.0 6.0 7.0 0.20 0.23 0.27 0.30 R@10 1.0 2.0 3.0 4.0 5.0 6.0 7.0 0.40 0.43 0.47 0.50 R@50 [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Comparison of the contextual disentanglement loss [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Multi-Type Context-Aware Conversational Recommender Systems via Mixture-of-Experts

    cs.CL 2025-04 conditional novelty 4.0 of 10

    A learned gating chair coordinates separate conversation, knowledge-graph, and review experts, and the paper reports improved movie recommendation accuracy and response diversity on ReDial and INSPIRED.

Reference graph

Works this paper leans on

56 extracted references · 47 canonical work pages · cited by 1 Pith paper

  1. [1]

    Guojia An, Jing Sun, Yuhan Yang, and Fuming Sun. 2024. Enhancing Collabora- tive Information with Contrastive Learning for Session-based Recommendation. Information Processing and Management 61, 4 (2024), 103738

    work page 2024

  2. [2]

    Qibin Chen, Junyang Lin, Yichang Zhang, Ming Ding, Yukuo Cen, Hongxia Yang, and Jie Tang. 2019. Towards Knowledge-Based Recommender Dialog System. arXiv preprint arXiv:1908.05391 (2019)

    arXiv 2019

  3. [3]

    Zhendong Chu, Hongning Wang, Yun Xiao, Bo Long, and Lingfei Wu. 2023. Meta Policy Learning for Cold-Start Conversational recommendation. InProceedings of the 16th ACM International Conference on Web Search and Data Mining . 222–230

    work page 2023

  4. [4]

    Huy Dao, Yang Deng, Dung D Le, and Lizi Liao. 2024. Broadening the View: Demonstration-augmented Prompt Learning for Conversational Recommenda- tion. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval . 785–795

    work page 2024

  5. [5]

    Yang Deng, Yaliang Li, Fei Sun, Bolin Ding, and Wai Lam. 2021. Unified Con- versational Recommendation Policy Learning via Graph-based Reinforcement Learning. In Proceedings of the 44th International ACM SIGIR Conference on Re- search and Development in Information Retrieval . 1431–1441

    work page 2021

  6. [6]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. In Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies . 4171–4186

    work page 2019

  7. [7]

    Shirley Anugrah Hayati, Dongyeop Kang, Qingxiaoyang Zhu, Weiyan Shi, and Zhou Yu. 2020. INSPIRED: Toward Sociable Recommendation Dialog Systems. In Proceedings of the Conference on Empirical Methods in Natural Language Processing. 8142–8152

    work page 2020

  8. [8]

    Dongxiao He, Jinghan Zhang, Xiaobao Wang, Meng Ge, Zhiyong Feng, Longbiao Wang, and Xiaoke Ma. 2024. TUT4CRS: Time-aware User-preference Tracking for Conversational Recommendation System. In Proceedings of the 32nd ACM International Conference on Multimedia . 5856–5864

    work page 2024

Show all 56 references
  1. [9]

    Taeho Kim, Juwon Yu, Won-Yong Shin, Hyunyoung Lee, Ji-hui Im, and Sang- Wook Kim. 2023. LATTE: A Framework for Learning Item-features to Make a Domain-Expert for Effective Conversational Recommendation. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery an...

    2023

  2. [10]

    Yoon Kim. 2014. Convolutional Neural Networks for Sentence Classification. In Proceedings of the 22nd Conference on Empirical Methods in Natural Language Processing. 1746–1751

    2014

  3. [11]

    Wenqiang Lei, Xiangnan He, Yisong Miao, Qingyun Wu, Richang Hong, Min-Yen Kan, and Tat-Seng Chua. 2020. Estimation-Action-Reflection: Towards Deep Interaction Between Conversational and Recommender Systems. In Proceedings of the 13th International Conference on Web Search and ...

    2020

  4. [12]

    Wenqiang Lei, Gangyi Zhang, Xiangnan He, Yisong Miao, Xiang Wang, Liang Chen, and Tat-Seng Chua. 2020. Interactive Path Reasoning on Graph for Conver- sational Recommendation. In Proceedings of the 26th International ACM SIGKDD Conference on Knowledge Discovery and Data Mining...

    2020

  5. [13]

    Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Veselin Stoyanov, and Luke Zettlemoyer. 2020. BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension. In Proceedings of the 58t...

    2020

  6. [14]

    Bobo Li, Hao Fei, Fei Li, Shengqiong Wu, Lizi Liao, Yinwei Wei, Tat-seng Chua, and Donghong Ji. 2025. Revisiting Conversation Discourse for Dialogue Disen- tanglement. Transactions on Information Systems 43, 1 (2025), 1–34

    2025

  7. [15]

    Raymond Li, Samira Ebrahimi Kahou, Hannes Schulz, Vincent Michalski, Laurent Charlin, and Chris Pal. 2018. Towards Deep Conversational Recommendations. In Proceedings of the 32nd International Conference on Neural Information Processing Systems. 9748–9758

    2018

  8. [16]

    Shijun Li, Wenqiang Lei, Qingyun Wu, Xiangnan He, Peng Jiang, and Tat-Seng Chua. 2021. Seamlessly unifying attributes and items: Conversational recom- mendation for cold-start users. Transactions on Information Systems 39, 4 (2021), 1–29

    2021

  9. [17]

    Wendi Li, Wei Wei, Xiaoye Qu, Xian-Ling Mao, Ye Yuan, Wenfeng Xie, and Dan- gyang Chen. 2023. TREA: Tree-Structure Reasoning Schema for Conversational Recommendation. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics. 2970–2982

    2023

  10. [18]

    Xinshu Li, Mingming Gong, and Lina Yao. 2024. Self-Distilled Disentangled Learning for Counterfactual Prediction. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining . 1667–1678

    2024

  11. [19]

    Dongding Lin, Jian Wang, and Wenjie Li. 2023. COLA: Improving Conversational Recommender Systems by Collaborative Augmentation. In Proceedings of the AAAI Conference on Artificial Intelligence . 4462–4470

    2023

  12. [20]

    Hui Liu, Zhan Shi, and Xiaodan Zhu. 2021. Unsupervised conversation disentan- glement through co-training. arXiv preprint arXiv:2109.03199 (2021)

    2021 arXiv

  13. [21]

    Yinhan Liu. 2019. RoBERTa: A Robustly Optimized BERT Pretraining Approach. arXiv preprint arXiv:1907.11692 364 (2019)

    2019 arXiv

  14. [22]

    Ilya Loshchilov and Frank Hutter. 2019. Decoupled Weight Decay Regularization. In Proceedings of the 7th International Conference on Learning Representations

    2019

  15. [23]

    Heli Ma, Jie Zou, Mohammad Aliannejadi, Evangelos Kanoulas, Yi Bin, and Yang Yang. 2024. Ask or Recommend: An Empirical Study on Conversational Product Search. In Proceedings of the 33rd ACM International Conference on Information and Knowledge Management. 3927–3931

    2024

  16. [24]

    Kong Menglin, Jia Wang, Yushan Pan, Haiyang Zhang, and Muzhou Hou. 2024. C2DR: Robust Cross-Domain Recommendation based on Causal Disentanglement. In Proceedings of the 17th ACM International Conference on Web Search and Data Mining. 341–349

    2024

  17. [25]

    Mingjie Qian, Yongsen Zheng, Jinghui Qin, and Liang Lin. 2023. HutCRS: Hi- erarchical User-Interest Tracking for Conversational Recommender System. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. 10281–10290

    2023

  18. [26]

    Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever, et al. 2019. Language Models are Unsupervised Multitask Learners. OpenAI blog 1, 8 (2019), 9

    2019

  19. [27]

    Xubin Ren, Lianghao Xia, Jiashu Zhao, Dawei Yin, and Chao Huang. 2023. Dis- entangled Contrastive Collaborative Filtering. In Proceedings of the 46th Inter- national ACM SIGIR Conference on Research and Development in Information Retrieval. 1137–1146

    2023

  20. [28]

    Xuhui Ren, Hongzhi Yin, Tong Chen, Hao Wang, Zi Huang, and Kai Zheng. 2021. Learning to Ask Appropriate Questions in Conversational Recommendation. In Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval . 808–817

    2021

  21. [29]

    Michael Schlichtkrull, Thomas N Kipf, Peter Bloem, Rianne Van Den Berg, Ivan Titov, and Max Welling. 2018. Modeling Relational Data with Graph Convolu- tional Networks. In Proceedings of the 15th International Conference on Semantic Web. 593–607

    2018

  22. [30]

    Iulian Serban, Alessandro Sordoni, Ryan Lowe, Laurent Charlin, Joelle Pineau, Aaron Courville, and Yoshua Bengio. 2017. A Hierarchical Latent Variable Encoder-Decoder Model for Generating Dialogues. In Proceedings of the AAAI conference on artificial intelligence , Vol. 31

    2017

  23. [31]

    Yueming Sun and Yi Zhang. 2018. Conversational Recommender System. In Proceedings of the 41st International ACM SIGIR Conference on Research and De- velopment in Information Retrieval . 235–244

    2018

  24. [32]

    Quan Tu, Shen Gao, Yanran Li, Jianwei Cui, Bin Wang, and Rui Yan. 2022. Conver- sational Recommendation via Hierarchical Information Modeling. In Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval. 2201–2205

    2022

  25. [33]

    Xiaolei Wang, Kun Zhou, Xinyu Tang, Wayne Xin Zhao, Fan Pan, Zhao Cao, and Ji-Rong Wen. 2023. Improving Conversational Recommendation Systems via Counterfactual Data Simulation. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining . 2398–2408

    2023

  26. [34]

    Xiaolei Wang, Kun Zhou, Ji-Rong Wen, and Wayne Xin Zhao. 2022. Towards Unified Conversational Recommender Systems via Knowledge-Enhanced Prompt Learning. In Proceedings of the 28th International ACM SIGKDD Conference on Knowledge Discovery and Data Mining . 1929–1937

    2022

  27. [35]

    Yibiao Wei, Jie Zou, Weikang Guo, Guoqing Wang, Xing Xu, and Yang Yang

  28. [36]

    Mingshi Yan, Fan Liu, Jing Sun, Fuming Sun, Zhiyong Cheng, and Yahong Han

  29. [37]

    Lu Zhang, Chen Li, Yu Lei, Zhu Sun, and Guanfeng Liu. 2024. An Empirical Analysis on Multi-turn Conversational Recommender Systems. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval. 841–851

    2024

  30. [38]

    Xiaotong Zhang, Xuefang Jia, Han Liu, Xinyue Liu, and Xianchao Zhang. 2024. A Goal Interaction Graph Planning Framework for Conversational Recommenda- tion. In Proceedings of the 38th AAAI Conference on Artificial Intelligence , Vol. 38. 19578–19587

    2024

  31. [39]

    Xiaoyu Zhang, Ruobing Xie, Yougang Lyu, Xin Xin, Pengjie Ren, Mingfei Liang, Bo Zhang, Zhanhui Kang, Maarten de Rijke, and Zhaochun Ren. 2024. Towards Empathetic Conversational Recommender Systems. In Proceedings of the 18th ACM Conference on Recommender Systems . 84–93

    2024

  32. [40]

    Xiaoyu Zhang, Xin Xin, Dongdong Li, Wenxuan Liu, Pengjie Ren, Zhumin Chen, Jun Ma, and Zhaochun Ren. 2023. Variational Reasoning over Incomplete Knowl- edge Graphs for Conversational Recommendation. InProceedings of the 16th ACM International Conference on Web Search and Data ...

    2023

  33. [41]

    Xiaokun Zhang, Bo Xu, Zhaochun Ren, Xiaochen Wang, Hongfei Lin, and Fen- glong Ma. 2024. Disentangling ID and Modality Effects for Session-based Rec- ommendation. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval...

    2024

  34. [42]

    Yizhe Zhang, Siqi Sun, Michel Galley, Yen-Chun Chen, Chris Brockett, Xiang Gao, Jianfeng Gao, Jingjing Liu, and Bill Dolan. 2020. DialoGPT : Large-Scale Beyond Whole Dialogue Modeling: Contextual Disentanglement for Conversational Recommendation SIGIR 2025, July 13–18, 2025, P...

    2020

  35. [43]

    Sen Zhao, Wei Wei, Ding Zou, and Xianling Mao. 2022. Multi-view Intent Disentangle Graph Networks for Bundle Recommendation. In Proceedings of the AAAI Conference on Artificial Intelligence , Vol. 36. 4379–4387

    2022

  36. [44]

    Yu Zheng, Chen Gao, Jianxin Chang, Yanan Niu, Yang Song, Depeng Jin, and Yong Li. 2022. Disentangling Long and Short-Term Interests for Recommendation. In Proceedings of the ACM Web Conference . 2256–2267

    2022

  37. [45]

    Yu Zheng, Chen Gao, Xiang Li, Xiangnan He, Yong Li, and Depeng Jin. 2021. Disentangling User Interest and Conformity for Recommendation with Causal Embedding. In Proceedings of the Web Conference . 2980–2991

    2021

  38. [46]

    Kun Zhou, Wayne Xin Zhao, Shuqing Bian, Yuanhang Zhou, Ji-Rong Wen, and Jingsong Yu. 2020. Improving Conversational Recommender Systems via Knowl- edge Graph based Semantic Fusion. In Proceedings of the 26th International ACM SIGKDD Conference on Knowledge Discovery and Data M...

    2020

  39. [47]

    Kun Zhou, Yuanhang Zhou, Wayne Xin Zhao, Xiaoke Wang, and Ji-Rong Wen

  40. [48]

    Yuanhang Zhou, Kun Zhou, Wayne Xin Zhao, Cheng Wang, Peng Jiang, and He Hu. 2022. C 2-CRS: Coarse-to-Fine Contrastive Learning for Conversational Recommender System. In Proceedings of the 15th ACM International Conference on Web Search and Data Mining . 1488–1496

    2022

  41. [49]

    Lixi Zhu, Xiaowen Huang, and Jitao Sang. 2024. How Reliable is Your Simu- lator? Analysis on the Limitations of Current LLM-based User Simulators for Conversational Recommendation. In Companion Proceedings of the ACM on Web Conference. 1726–1732

    2024

  42. [50]

    Jie Zou, Yifan Chen, and Evangelos Kanoulas. 2020. Towards Question-based Rec- ommender System. In Proceedings of the 43rd international ACM SIGIR Conference on Research and Development in Information Retrieval . 881–890

    2020

  43. [51]

    Jie Zou, Jimmy Huang, Zhaochun Ren, and Evangelos Kanoulas. 2022. Learning to ask: Conversational product search via representation learning. Transactions on Information Systems 41, 2 (2022), 1–27

    2022

  44. [52]

    Jie Zou, Evangelos Kanoulas, Pengjie Ren, Zhaochun Ren, Aixin Sun, and Cheng Long. 2022. Improving Conversational Recommender Systems via Transformer- based Sequential Modelling. In Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Infor...

    2022

  45. [53]

    Jie Zou, Aixin Sun, Cheng Long, and Evangelos Kanoulas. 2024. Knowledge- Enhanced Conversational Recommendation via Transformer-Based Sequential Modeling. Transactions on Information Systems 42, 6 (2024), 1–27

    2024

  46. [2020]

    In Proceed- ings of the 28th International Conference on Computational Linguistics

    Towards Topic-Guided Conversational Recommender System. In Proceed- ings of the 28th International Conference on Computational Linguistics. 4128–4139

  47. [2024]

    In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval

    Behavior-Contextualized Item Preference Modeling for Multi-Behavior Recommendation. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval . 946–955

  48. [2025]

    arXiv preprint arXiv:2504.10921 (2025)

    MSCRS: Multi-modal Semantic Graph Prompt Learning Framework for Conversational Recommender Systems. arXiv preprint arXiv:2504.10921 (2025)

    2025 arXiv

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.