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

Cheng Lin; Heng Tao Shen; Jie Zou; Jiwei Wei; Weikang Guo; Yang Yang; Zheng Wang

REVIEW 4 major objections 6 minor 72 references

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

T0 review · 4 major / 6 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 MCCRS fuses conversation history, a knowledge graph, and item reviews through three experts coordinated by a ChairBot, reporting higher Recall@1/10/50 and better response diversity than state-of-the-art baselines on ReDial and INSPIRED.

desk verdict A reasonable, incremental fusion architecture for conversational recommendation, with a load-bearing gating equation that is wrong as written and evaluation choices that need tightening. read the letter →

arxiv 2504.13655 v2 pith:H3GNZFAA submitted 2025-04-18 cs.CL cs.AIcs.IR

Jie Zou , Cheng Lin , Weikang Guo , Zheng Wang , Jiwei Wei , Yang Yang , Heng Tao Shen This is my paper · ORCID

classification cs.CLcs.AIcs.IR

keywords conversationalrecommendersystemmixture-of-expertsChairBotknowledgegraphitemreviewsconversationhistoryReDialINSPIRED

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 tries to establish that a conversational recommender system improves when three heterogeneous sources—conversation history, a structured knowledge graph, and item reviews—are each handled by a dedicated expert and then blended by a learned ChairBot, rather than aligned by contrastive learning or read through a single source. The resulting model, MCCRS, reports higher Recall@1, Recall@10, and Recall@50 than state-of-the-art baselines on the ReDial and INSPIRED datasets, and its generated responses score higher on distinctness and human-rated fluency and informativeness. A sympathetic reader would take the contribution to be evidence that per-source specialization plus learned gating is a workable answer to the data-heterogeneity and missing-alignment problems of multi-type fusion in conversational recommendation.

What carries the argument

The load-bearing mechanism is the ChairBot's gating identity: three separately trained experts each output an item probability, and the ChairBot combines them as a weighted sum. The conversation expert is a transformer with a Cloze-style training objective that randomly masks and predicts items; the graph expert is a relational graph convolutional network (R-GCN) over DBpedia entities with a self-attention user representation; the review expert encodes review sentences with a transformer and sentence-level self-attention. The ChairBot forms per-expert representations $h^b = h^b_i \oplus p^b_i$, maps them through a linear layer to $\beta_b$, normalizes by $\beta_C+\beta_G+\beta_R$ to get $\lambda_b$, and outputs $P_{\mathrm{rec}}(i)=\lambda_C P_C(i)+\lambda_G P_G(i)+\lambda_R P_R(i)$ (Eq. 12). This identity carries the argument because, if it works, the model needs no contrastive alignment between structured and unstructured sources; it only needs to learn how much to trust each expert for each candidate item.

What would settle it

Inspect the trained ChairBot's $\lambda_b$ values on held-out conversations: if any weight is negative or any $P_{\mathrm{rec}}(i)$ from Eq. (12) leaves $[0,1]$, the 'normalized importance score' is not a valid probability fusion. Retraining with a softmax-constrained gating in place of Eq. (11) and checking whether the reported Recall@1/10/50 gains over C2-CRS and VRICR persist would directly test whether ChairBot gating itself, rather than extra parameters or unconstrained scores, drives the improvement.

Watch

Extended reading notes

Core claim

The central claim is that a mixture-of-experts design, with one expert per contextual source, outperforms both single-source models and contrastively fused multi-source models. MCCRS consists of a conversation expert (a transformer reading the extracted entity sequence), a graph expert (an R-GCN encoding DBpedia entities), and a review expert (a transformer with self-attention over item reviews). Each expert trains separately and emits an item probability, $P_C(i)$, $P_G(i)$, $P_R(i)$; the ChairBot concatenates each expert's hidden state and probability, maps the result through a linear layer to importance scores $\beta_b$, normalizes those scores, and computes $P_{\mathrm{rec}}(i)=\lambda_C P_C(i)+\lambda_G P_G(i)+\lambda_R P_R(i)$. With this fused score, MCCRS reports Recall@1 of 0.057 (best baseline 0.054), Recall@10 of 0.250 (best baseline 0.244), and Recall@50 of 0.473 (best baseline 0.407) on ReDial, along with corresponding gains on INSPIRED, higher Distinct-n response scores, and better human ratings. The paper interprets these results as showing that the ChairBot's gating breaks the bottleneck of relying on any single contextual information.

Load-bearing premise

The load-bearing assumption is that the ChairBot's learned weights combine three expert probabilities into a valid probability; nothing in the equations forces the weights non-negative, and the experts' scores are assumed to live on comparable scales.

Editorial extensions

If this is right

If MCCRS is right, conversational recommenders can fuse structured and unstructured sources without contrastive alignment, sidestepping the requirement that all sources share the same entries.
Specialized experts with a ChairBot make the system more traceable: when a recommendation is good or bad, one can attribute it to the conversation, graph, or review expert and adjust that expert alone.
The ablation results imply the knowledge-graph expert carries the most weight, so future systems should keep structured knowledge even when they add review or conversation models.
The architecture extends by adding new experts for new external sources, so multi-source conversational recommendation research can grow incrementally rather than retraining a single fused model.
Because the response generator draws each expert's embeddings through cross-attention, improvements in the recommender propagate directly into more informative and diverse generated responses.

Reading between the lines

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

Editorial inference: the gating design predicts a scaling property—adding further specialist experts (for instance, sentiment or multimodal signals) should keep helping only if the ChairBot's weights stay well calibrated; if the gate collapses to constant weights, the mixture adds parameters without adding signal.
Editorial inference: the paper's case against contrastive fusion implies a testable comparison on datasets where conversation entities and review texts deliberately do not overlap; a calibrated MCCRS should degrade less than contrastive baselines that need shared entries.
Editorial inference: Section 5 concedes that entity linking from conversations to DBpedia may be imperfect, so the graph expert's large ablation contribution should be revalidated under noisy or incomplete entity annotations.
Editorial inference: the reported sensitivity to hidden dimension (best at 32) and mask probability (best at 0.4) suggests the advantage may be configuration-dependent; adaptive schedules for these hyperparameters would test how robust the gating advantage is.

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.

Desk Editor's Note

read the letter

Quick take: MCCRS is a plausible, incremental fusion model for conversational recommendation. The new piece is a learned ChairBot that gates three separately trained experts—a conversation transformer, an R-GCN on DBpedia, and a review encoder—and feeds their representations into the decoder. The components are mostly borrowed, including the conversation expert from the authors' own prior work, but the specific combination is new and the paper shows it works on ReDial and INSPIRED.

What the paper does well: the architecture is easy to follow, the ablations are informative, and the recommendation gains over C2-CRS and VRICR are consistent across both datasets. The review expert helps only marginally, which the authors acknowledge; that is honest. The response-diversity numbers are plausible given the multi-source cross-attention, though I would not over-read them.

The soft spots are real. First, Eq. 11–12, the core of the ChairBot, is wrong as written. β_b is an unconstrained linear-layer output; dividing by the sum gives λ_b that can be negative or undefined, so P_rec(i) can leave [0,1] and the cross-entropy objective in Eq. 5 is not well-defined. This is almost certainly a missing softmax or positive activation, but the manuscript does not say so, and with no code release, the stated algorithm cannot be reproduced. The stress-test note is right on this point.

Second, the parameter sensitivity analysis selects the mask proportion and hidden size by looking at test-set Recall (Tables 9 and 10), which inflates the reported numbers. The main tables have no error bars and only a terse significance footnote. Third, the human evaluation uses three annotators with no agreement metric. Fourth, all baselines are pre-LLM; the authors justify this, but the SOTA claim should be scoped accordingly.

None of this sinks the central idea. The gating is a natural extension, and the ablations suggest the fusion is doing work. The flaws are in presentation and validation details: the equations need fixing, the hyperparameter procedure needs to move to validation, and code or at least detailed experiment setup should be released.

My recommendation: send it to peer review. A good referee can push for the gating fix and tighter evaluation, and the result would be a solid subfield paper. It is not a desk reject.

Referee Report

4 major / 6 minor

Summary. The paper proposes MCCRS, a conversational recommender system that fuses three types of contextual information—conversation history, a DBpedia knowledge graph, and item reviews—through a mixture-of-experts architecture. Each of the three experts (conversation, graph, and review) is trained with a cross-entropy objective on the recommendation task, and a ChairBot combines the experts' probability outputs via a learned weighted sum. The fused representation is then used both for item recommendation and, through cross-attention in a transformer decoder, for response generation. Experiments on ReDial and INSPIRED compare MCCRS against several classical and recent conversational recommender baselines, with ablation and parameter sensitivity studies on ReDial. The authors report significant improvements in recommendation recall and response diversity, as well as human evaluation of fluency and informativeness.

Significance. If the implementation matches the description, the paper shows a clean and extensible fusion mechanism for multi-source context in conversational recommender systems, with ablations indicating that all three experts contribute to the final performance. The use of a ChairBot that linearly combines expert probabilities is a simple yet plausible alternative to contrastive-learning-based fusion. However, the validity of the central contribution is currently weakened by an under-specified gating equation that can produce invalid probabilities, by the absence of variance measures and detailed significance tests for the reported improvements, and by unclear hyperparameter selection protocol that may involve tuning on the test set. The architecture is nonetheless reasonable and the empirical claims are potentially reproducible after these issues are addressed.

major comments (4)

[§3.5, Eqs. (11)–(12)] The gating equation as written does not define a valid fused probability distribution. In Eq. (11), β_b = MLP(h_b) is described as a linear layer with no non-negativity constraint, and λ_b = β_b / (β_C + β_G + β_R) can therefore be negative or undefined if the denominator equals zero. Consequently, P_rec(i) in Eq. (12) can lie outside [0,1] or be undefined, and the cross-entropy loss in Eq. (5) would be undefined for any item with P_rec(i) ≤ 0. Please replace the normalization in Eq. (11) with a softmax over the β_b scores (and add a small stability constant to the denominator), or otherwise provide an explicit non-negativity constraint. This repair is necessary for the reported training procedure to be well-defined and for the results to be reproducible from the specification.
[§4.4, Tables 9–10] The mask proportion and hidden dimensionality are selected based on their effect on ReDial recommendation metrics in Tables 9–10, but the paper does not state whether these 'optimal' values were chosen on the validation split or directly on the test set. Since the final reported test results in Tables 3–4 depend on these hyperparameters, tuning on the test set would bias the reported improvements and break comparability with baselines. Please clarify that all hyperparameter choices were made using the validation split alone, or re-run the selection on validation and update the tables accordingly.
[§4.2, Tables 3–6] The significance claims (indicated by '*' and attributed to a Fisher random test with p < 0.05) are reported without any measure of variance, such as standard deviations or confidence intervals across runs or conversations, and the test procedure is not described (what is randomized, how many permutations, and what the unit of randomization is). For small differences such as Recall@1 = 0.057 vs 0.053 in Table 3, a significance claim without variance or a detailed test description is not verifiable. Please report means and standard deviations over multiple random seeds and describe the significance test in enough detail to allow reproduction.
[§4.2.2, Table 7] The human evaluation reports only mean scores from three annotators for fluency and informativeness. There is no inter-annotator agreement measure (e.g., Fleiss' kappa), no per-model or per-item variance, and no significance test. The observed differences (e.g., 1.66 vs 1.58 for fluency) may be within the noise of a 0–2 scale with three annotators, so the claim that MCCRS generates significantly more fluent and informative responses is not sufficiently supported. Please provide agreement statistics, variance information, and an appropriate significance test, or temper the corresponding claim.

minor comments (6)

[§3.2, Eq. (3)] The parenthesis in Eq. (3) is unbalanced; the equation should read Hn = LayerNorm(Hn + Dropout(sublayer(Hn))).
[§4.1.3 and §4.4] The hyperparameters mask proportion and hidden dimensionality are said to be discussed in Section 4.4, but Section 4.4 only reports experiments on ReDial; please specify the values used for the INSPIRED experiments as well.
[Table 6] The Distinct-n values in Table 6 are larger than 1 (e.g., Transformer 1.020, MCCRS 2.584); if these are percentages multiplied by 100, please state this in the caption or table notes.
[§4.1.1] Footnote 3 is placed after 'we retrieve reviews for movies from IMDb' but the footnote text points to dbpedia.org; this appears to be a mismatched citation.
[§3.2] The description of the transformer as using 'a matrix of hidden size × vocabulary size in place of the usual positional encoding' is unclear; please clarify what this matrix is and how it is combined with the sequence embedding in Eq. (1).
[General] The paper does not mention whether code and trained models will be released; given the ambiguity in the gating equations, providing code or pseudocode would substantially improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the method is an empirical supervised-learning architecture, each expert is fully specified by equations in the paper, and the reported gains are measured against external benchmarks.

full rationale

The paper proposes MCCRS, an architecture combining three independently trained experts (a conversation Transformer, an R-GCN graph encoder, and a review self-attention encoder) and a ChairBot that computes a weighted sum of expert probabilities. There is no derivation chain in which a predicted quantity is defined from a fitted parameter: each expert is trained with a cross-entropy loss on the same recommendation targets, and the ChairBot's output is a learned linear combination rather than a restatement of any single input. The only self-citation of note is the conversation expert's use of the authors' earlier transformer-based sequential model [2]; however, the expert is fully specified in Eqs. (1)-(4), so the citation is a reference to a published component rather than load-bearing circular evidence. The footnote-2 statement that separate expert training was chosen because it performed better in experiments is a model-selection protocol concern, not a circular reduction, because the final reported predictions are evaluated on held-out test conversations and are not equal by construction to the training objective. The manuscript's own limitations about entity noise, missing reviews, sentiment, and computational cost do not assert any circular step. Eq. (12)'s normalization issue, where unconstrained beta values can make lambda negative and the weighted sum fall outside [0,1], is a correctness and reproducibility risk rather than circularity. Therefore the paper is self-contained against external benchmarks and deserves a score of 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim rests on supervised training of three experts and a gating module; no new physical or mathematical content is assumed beyond standard ML machinery. The main assumptions are domain-level: entity linking accuracy, review alignment, and comparability of expert outputs.

free parameters (3)

Expert gating weights lambda_C, lambda_G, lambda_R = Not reported
Eqs. 11-12: the ChairBot learns scalar weights to combine expert probabilities; this is the core fusion mechanism and is fit to the recommendation objective.
Mask proportion = 0.4
Chosen as optimal on ReDial test set in Table 9; no validation-based selection is shown.
Hidden dimensionality = 32
Chosen as best on ReDial test set in Table 10; no validation-based selection is shown.

assumptions (4)

domain assumption Entity linking via TagMe maps conversation mentions to DBpedia entities accurately enough for preference modeling.
Invoked in Section 3.3 and acknowledged in Section 5 as a limitation; if linking is noisy, the graph expert's representations degrade.
domain assumption Item reviews retrieved from IMDb are aligned with the movies and carry useful preference signal.
Section 4.1.1 states reviews are retrieved from IMDb similar to Zhou et al. [14]; missing reviews are filled with mean representations, which assumes missingness is ignorable.
ad hoc to paper The outputs of the three experts are comparable and can be linearly combined into a valid fused distribution via Eq. 12.
Section 3.5 introduces unconstrained beta outputs and calls lambda normalized; no derivation guarantees non-negative weights or a valid probability.
standard math Standard transformer, R-GCN, and attention machinery works as expected.
Sections 3.2-3.4 rely on established architectures without modification.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Multi-Type Context-Aware Conversational Recommender Systems via Mixture-of-Experts." pith.science (2026). https://pith.science/paper/H3GNZFAA

@misc{pith2026250413655,
  author       = {Pith},
  title        = {Pith review of: Multi-Type Context-Aware Conversational Recommender Systems via Mixture-of-Experts},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/H3GNZFAA}},
  note         = {Machine review of arXiv:2504.13655}
}

read the original abstract

Conversational recommender systems enable natural language conversations and thus lead to a more engaging and effective recommendation scenario. As the conversations for recommender systems usually contain limited contextual information, many existing conversational recommender systems incorporate external sources to enrich the contextual information. However, how to combine different types of contextual information is still a challenge. In this paper, we propose a multi-type context-aware conversational recommender system, called MCCRS, effectively fusing multi-type contextual information via mixture-of-experts to improve conversational recommender systems. MCCRS incorporates both structured information and unstructured information, including the structured knowledge graph, unstructured conversation history, and unstructured item reviews. It consists of several experts, with each expert specialized in a particular domain (i.e., one specific contextual information). Multiple experts are then coordinated by a ChairBot to generate the final results. Our proposed MCCRS model takes advantage of different contextual information and the specialization of different experts followed by a ChairBot breaks the model bottleneck on a single contextual information. Experimental results demonstrate that our proposed MCCRS method achieves significantly higher performance compared to existing baselines.

Figures

Figures reproduced from arXiv: 2504.13655 by the authors.

**Figure 2.** Illustration of an example graph. where W is a learnable transformation matrix, and b denotes the trainable bias matrix. On top of the produced probability of items, we use a crossentropy loss to optimize the parameters of the conversation expert. Lrec = − X j=1 X i∈I yij · log P(i), (5) where j is the conversation index, and yij indicates the ground-truth label of items. After the training, we collect both the pre… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

72 extracted references · 60 canonical work pages

[1]

C. Gao, W. Lei, X. He, M. de Rijke, T.-S. Chua, Advances and challenges in conversational recommender systems: A survey, AI Open 2 (2021) 100–126

work page 2021
[2]

J. Zou, E. Kanoulas, P. Ren, Z. Ren, A. Sun, C. Long, Improving conversational recommender systems via transformer-based sequential modelling, in: Proceedings of the 45th International ACM SIGIR Con- ference on Research and Development in Information Retrieval, 2022, pp. 2319–2324

work page 2022
[3]

Q. Chen, J. Lin, Y. Zhang, M. Ding, Y. Cen, H. Yang, J. Tang, To- wards knowledge-based recommender dialog system, in: Proceedings of 23 the 2019 Conference on Empirical Methods in Natural Language Pro- cessing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), 2019, pp. 1803–1813

work page 2019
[4]

K. Zhou, W. X. Zhao, S. Bian, Y. Zhou, J.-R. Wen, J. Yu, Improving conversational recommender systems via knowledge graph based seman- tic fusion, in: Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2020, pp. 1006– 1014

work page 2020
[5]

Sarkar, K

R. Sarkar, K. Goswami, M. Arcan, J. P. McCrae, Suggest me a movie for tonight: Leveraging knowledge graphs for conversational recommen- dation, in: Proceedings of the 28th International Conference on Com- putational Linguistics, 2020, pp. 4179–4189

work page 2020
[6]

W. Ma, R. Takanobu, M. Huang, CR-walker: Tree-structured graph rea- soning and dialog acts for conversational recommendation, in: Proceed- ings of the 2021 Conference on Empirical Methods in Natural Language Processing, 2021, pp. 1839–1851

work page 2021
[7]

K. Zhou, Y. Zhou, W. X. Zhao, X. Wang, J.-R. Wen, Towards topic- guided conversational recommender system, in: Proceedings of the 28th International Conference on Computational Linguistics, 2020, pp. 4128– 4139

work page 2020
[8]

X. Wang, Y. Li, J. Liu, Conversation recommender system based on knowledge graph and time-series feature, in: The 5th International Conference on Computer Science and Application Engineering, 2021, pp. 1–5

work page 2021

Show all 72 references

[9]

L. Wang, H. Hu, L. Sha, C. Xu, D. Jiang, K.-F. Wong, Recindial: A unified framework for conversational recommendation with pretrained language models, in: Proceedings of the 2nd Conference of the Asia- Pacific Chapter of the Association for Computational Linguistics and the 12...

2022
[10]

K. Chen, S. Sun, Knowledge-based conversational recommender systems enhanced by dialogue policy learning, in: The 10th International Joint Conference on Knowledge Graphs, 2021, pp. 10–18. 24

2021
[11]

B. Yang, C. Han, Y. Li, L. Zuo, Z. Yu, Improving conversational recom- mendation systems’ quality with context-aware item meta information, arXiv preprint arXiv:2112.08140 (2021)

2021 arXiv
[12]

Zhang, Y

T. Zhang, Y. Liu, P. Zhong, C. Zhang, H. Wang, C. Miao, Ke- crs: Towards knowledge-enriched conversational recommendation sys- tem, arXiv preprint arXiv:2105.08261 (2021)

2021 arXiv
[13]

Y. Lu, J. Bao, Y. Song, Z. Ma, S. Cui, Y. Wu, X. He, Revcore: Review-augmented conversational recommendation, arXiv preprint arXiv:2106.00957 (2021)

2021 arXiv
[14]

Y. Zhou, K. Zhou, W. X. Zhao, C. Wang, P. Jiang, H. Hu, C 2-crs: Coarse-to-fine contrastive learning for conversational recommender sys- tem, in: Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining, 2022, pp. 1488–1496

2022
[15]

Z. Ren, Z. Tian, D. Li, P. Ren, L. Yang, X. Xin, H. Liang, M. de Ri- jke, Z. Chen, Variational reasoning about user preferences for conver- sational recommendation, in: Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrie...

2022
[16]

H. Won, B. Oh, H. Yang, K.-H. Lee, Cross-modal contrastive learning for aspect-based recommendation, Information Fusion 99 (2023) 101858

2023
[17]

Jannach, A

D. Jannach, A. Manzoor, W. Cai, L. Chen, A survey on conversational recommender systems, ACM Computing Surveys (CSUR) 54 (2021) 1–36

2021
[18]

T. Shen, J. Li, M. R. Bouadjenek, Z. Mai, S. Sanner, Towards un- derstanding and mitigating unintended biases in language model-driven conversational recommendation, Information Processing & Management 60 (2023) 103139

2023
[19]

S. Li, R. Xie, Y. Zhu, F. Zhuang, Z. Tang, W. X. Zhao, Q. He, Self- supervised learning for conversational recommendation, Information Processing & Management 59 (2022) 103067. 25

2022
[20]

Zhang, X

X. Zhang, X. Jia, H. Liu, X. Liu, X. Zhang, A goal interaction graph planning framework for conversational recommendation, in: Proceed- ings of the AAAI Conference on Artificial Intelligence, volume 38, 2024, pp. 19578–19587

2024
[21]

D. He, J. Zhang, X. Wang, M. Ge, Z. Feng, L. Wang, X. Ma, Tut4crs: Time-aware user-preference tracking for conversational recommendation system, in: Proceedings of the 32nd ACM International Conference on Multimedia, 2024, pp. 5856–5864

2024
[22]

J. Zou, A. Sun, C. Long, E. Kanoulas, Knowledge-enhanced conver- sational recommendation via transformer-based sequential modeling, ACM Transactions on Information Systems 42 (2024) 1–27

2024
[23]

J. Zou, Y. Chen, E. Kanoulas, Towards question-based recommender systems, in: Proceedings of the 43rd International ACM SIGIR Con- ference on Research and Development in Information Retrieval, SIGIR ’20, 2020, pp. 881–890

2020
[24]

Zhang, R

X. Zhang, R. Xie, Y. Lyu, X. Xin, P. Ren, M. Liang, B. Zhang, Z. Kang, M. de Rijke, Z. Ren, Towards empathetic conversational recommender systems, in: Proceedings of the 18th ACM Conference on Recommender Systems, 2024, pp. 84–93

2024
[25]

Zhang, X

Y. Zhang, X. Chen, Q. Ai, L. Yang, W. B. Croft, Towards conversational search and recommendation: System ask, user respond, in: Proceedings of the 27th acm international conference on information and knowledge management, 2018, pp. 177–186

2018
[26]

Y. Sun, Y. Zhang, Conversational recommender system, in: The 41st international acm sigir conference on research & development in infor- mation retrieval, 2018, pp. 235–244

2018
[27]

Christakopoulou, A

K. Christakopoulou, A. Beutel, R. Li, S. Jain, E. H. Chi, Q&r: A two-stage approach toward interactive recommendation, in: Proceed- ings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2018, pp. 139–148

2018
[28]

W. Lei, X. He, Y. Miao, Q. Wu, R. Hong, M.-Y. Kan, T.-S. Chua, Estimation-action-reflection: Towards deep interaction between conver- 26 sational and recommender systems, in: Proceedings of the 13th Interna- tional Conference on Web Search and Data Mining, 2020, pp. 304–312

2020
[29]

W. Lei, G. Zhang, X. He, Y. Miao, X. Wang, L. Chen, T.-S. Chua, Interactive path reasoning on graph for conversational recommendation, in: Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, KDD ’20, 2020, pp. 2073–2083

2020
[30]

C. Hu, S. Huang, Y. Zhang, Y. Liu, Learning to infer user implicit pref- erence in conversational recommendation, in: Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2022, pp. 256–266

2022
[31]

Zhang, L

Y. Zhang, L. Wu, Q. Shen, Y. Pang, Z. Wei, F. Xu, B. Long, J. Pei, Multiple choice questions based multi-interest policy learning for conver- sational recommendation, in: Proceedings of the ACM Web Conference 2022, 2022, pp. 2153–2162

2022
[32]

Christakopoulou, F

K. Christakopoulou, F. Radlinski, K. Hofmann, Towards conversational recommender systems, in: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, ACM, 2016, pp. 815–824

2016
[33]

S. Li, W. Lei, Q. Wu, X. He, P. Jiang, T.-S. Chua, Seamlessly unify- ing attributes and items: Conversational recommendation for cold-start users, ACM Transactions on Information Systems (TOIS) 39 (2021) 1–29

2021
[34]

X. Dai, Z. Wang, J. Xie, X. Liu, J. C. Lui, Conversational recommen- dation with online learning and clustering on misspecified users, IEEE Transactions on Knowledge and Data Engineering (2024)

2024
[35]

J. Zou, J. Huang, Z. Ren, E. Kanoulas, Learning to ask: Conversa- tional product search via representation learning, ACM Transactions on Information Systems 41 (2022) 1–27

2022
[36]

Z. Liu, H. Wang, Z.-Y. Niu, H. Wu, W. Che, T. Liu, Towards conversa- tional recommendation over multi-type dialogs, in: Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 2020, pp. 1036–1049. 27

2020
[37]

L. Liao, R. Takanobu, Y. Ma, X. Yang, M. Huang, T.-S. Chua, Deep conversational recommender in travel, arXiv preprint arXiv:1907.00710 (2019)

2019 arXiv
[38]

Liang, H

Z. Liang, H. Hu, C. Xu, J. Miao, Y. He, Y. Chen, X. Geng, F. Liang, D. Jiang, Learning neural templates for recommender dialogue system, in: Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, 2021, pp. 7821–7833

2021
[39]

Y. Deng, W. Zhang, W. Xu, W. Lei, T.-S. Chua, W. Lam, A uni- fied multi-task learning framework for multi-goal conversational recom- mender systems, ACM Transactions on Information Systems (2022)

2022
[40]

Y. Pan, Y. Yin, F. Huang, Keyword-guided topic-oriented conversa- tional recommender system, in: 2022 International Joint Conference on Neural Networks (IJCNN), 2022, pp. 1–8

2022
[41]

R. Li, S. E. Kahou, H. Schulz, V. Michalski, L. Charlin, C. Pal, To- wards deep conversational recommendations, in: Advances in Neural Information Processing Systems, 2018, pp. 9748–9758

2018
[42]

S. A. Hayati, D. Kang, Q. Zhu, W. Shi, Z. Yu, Inspired: Toward sociable recommendation dialog systems, in: Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2020, pp. 8142–8152

2020
[43]

Zhang, Y

J. Zhang, Y. Yang, C. Chen, L. He, Z. Yu, Kers: A knowledge-enhanced framework for recommendation dialog systems with multiple subgoals, in: Findings of the Association for Computational Linguistics: EMNLP 2021, 2021, pp. 1092–1101

2021
[44]

X. Ren, T. Chen, Q. V. H. Nguyen, L. Cui, Z. Huang, H. Yin, Explicit knowledge graph reasoning for conversational recommendation, ACM Transactions on Intelligent Systems and Technology 15 (2024) 1–21

2024
[45]

W. Li, W. Wei, X. Qu, X.-L. Mao, Y. Yuan, W. Xie, D. Chen, Trea: Tree-structure reasoning schema for conversational recommendation, in: Proceedings of the 61st Annual Meeting of the Association for Compu- tational Linguistics (Volume 1: Long Papers), 2023, pp. 2970–2982. 28

2023
[46]

Z. Qiu, Y. Tao, S. Pan, A. W.-C. Liew, Knowledge graphs and pre- trained language models enhanced representation learning for conversa- tional recommender systems, IEEE Transactions on Neural Networks and Learning Systems (2024)

2024
[47]

Bizer, J

C. Bizer, J. Lehmann, G. Kobilarov, S. Auer, C. Becker, R. Cyganiak, S. Hellmann, Dbpedia-a crystallization point for the web of data, Jour- nal of web semantics 7 (2009) 154–165

2009
[48]

J. Zhu, C. Huang, P. De Meo, Dfmke: A dual fusion multi-modal knowledge graph embedding framework for entity alignment, Informa- tion Fusion 90 (2023) 111–119

2023
[49]

Speer, J

R. Speer, J. Chin, C. Havasi, Conceptnet 5.5: an open multilingual graph of general knowledge, in: Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence, 2017, pp. 4444–4451

2017
[50]

J. Zhou, B. Wang, R. He, Y. Hou, Crfr: Improving conversational recom- mender systems via flexible fragments reasoning on knowledge graphs, in: Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, 2021, pp. 4324–4334

2021
[51]

Zhang, X

X. Zhang, X. Xin, D. Li, W. Liu, P. Ren, Z. Chen, J. Ma, Z. Ren, Vari- ational reasoning over incomplete knowledge graphs for conversational recommendation, in: Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining, 2023, pp. 231–239

2023
[52]

X. Wang, K. Zhou, J.-R. Wen, W. X. Zhao, Towards unified conversa- tional recommender systems via knowledge-enhanced prompt learning, in: Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2022, pp. 1929–1937

2022
[53]

X. Wang, X. Tang, W. X. Zhao, J. Wang, J.-R. Wen, Rethinking the evaluation for conversational recommendation in the era of large lan- guage models, in: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, 2023, pp. 10052–10065

2023
[54]

Y. Xi, W. Liu, J. Lin, B. Chen, R. Tang, W. Zhang, Y. Yu, Memocrs: Memory-enhanced sequential conversational recommender systems with large language models, arXiv preprint arXiv:2407.04960 (2024). 29

2024 arXiv
[55]

K. D. Spurlock, C. Acun, E. Saka, O. Nasraoui, Chatgpt for conver- sational recommendation: Refining recommendations by reprompting with feedback, arXiv preprint arXiv:2401.03605 (2024)

2024 arXiv
[56]

Zhang, X

C. Zhang, X. Huang, J. An, S. Zou, Improving conversational recommender systems via multi-preference modelling and knowledge- enhanced, Knowledge-Based Systems 286 (2024) 111361

2024
[57]

H. Dao, Y. Deng, D. D. Le, L. Liao, Broadening the view: Demonstration-augmented prompt learning for conversational recom- mendation, in: Proceedings of the 47th International ACM SIGIR Con- ference on Research and Development in Information Retrieval, 2024, pp. 785–795

2024
[58]

Kemper, J

S. Kemper, J. Cui, K. Dicarlantonio, K. Lin, D. Tang, A. Korikov, S. Sanner, Retrieval-augmented conversational recommendation with prompt-based semi-structured natural language state tracking, in: Pro- ceedings of the 47th International ACM SIGIR Conference on Research and De...

2024
[59]

G. An, J. Zou, J. Wei, C. Zhang, F. Sun, Y. Yang, Beyond whole dialogue modeling: Contextual disentanglement for conversational rec- ommendation, arXiv preprint arXiv:2504.17427 (2025)

2025 arXiv
[60]

Y. Wei, J. Zou, W. Guo, G. Wang, X. Xu, Y. Yang, Mscrs: Multi- modal semantic graph prompt learning framework for conversational recommender systems, arXiv preprint arXiv:2504.10921 (2025)

2025 arXiv
[61]

S. Li, R. Xie, Y. Zhu, X. Ao, F. Zhuang, Q. He, User-centric conversa- tional recommendation with multi-aspect user modeling, in: Proceed- ings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2022, pp. 223–233

2022
[62]

Devlin, M.-W

J. Devlin, M.-W. Chang, K. Lee, K. Toutanova, Bert: Pre-training of deep bidirectional transformers for language understanding, arXiv preprint arXiv:1810.04805 (2018)

2018 arXiv
[63]

cloze procedure

W. L. Taylor, “cloze procedure”: A new tool for measuring readability, Journalism quarterly 30 (1953) 415–433. 30

1953
[64]

Hendrycks, K

D. Hendrycks, K. Gimpel, Bridging nonlinearities and stochastic regu- larizers with gaussian error linear units, CoRR, abs/1606.08415 3 (2016)

2016 arXiv
[65]

Lehmann, R

J. Lehmann, R. Isele, M. Jakob, A. Jentzsch, D. Kontokostas, P. N. Mendes, S. Hellmann, M. Morsey, P. Van Kleef, S. Auer, et al., Dbpedia– a large-scale, multilingual knowledge base extracted from wikipedia, Se- mantic web 6 (2015) 167–195

2015
[66]

Ferragina, U

P. Ferragina, U. Scaiella, Tagme: on-the-fly annotation of short text fragments (by wikipedia entities), in: Proceedings of the 19th ACM international conference on Information and knowledge management, 2010, pp. 1625–1628

2010
[67]

Schlichtkrull, T

M. Schlichtkrull, T. N. Kipf, P. Bloem, R. Van Den Berg, I. Titov, M. Welling, Modeling relational data with graph convolutional networks, in: The Semantic Web: 15th International Conference, ESWC 2018, Heraklion, Crete, Greece, June 3–7, 2018, Proceedings 15, 2018, pp. 593–607

2018
[68]

J. Pei, P. Ren, M. de Rijke, A modular task-oriented dialogue system using a neural mixture-of-experts (2019). arXiv:1907.05346

2019 arXiv
[69]

D. P. Kingma, J. Ba, Adam: A method for stochastic optimization, arXiv preprint arXiv:1412.6980 (2014)

2014 arXiv
[70]

Kim, Convolutional neural networks for sentence classification, arXiv preprint arXiv:1408.5882 (2014)

Y. Kim, Convolutional neural networks for sentence classification, arXiv preprint arXiv:1408.5882 (2014)

2014 arXiv
[71]

Vaswani, N

A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, I. Polosukhin, Attention is all you need, in: Advances in neural information processing systems, 2017, pp. 5998–6008

2017
[72]

C. Lin, H. Hu, J. Zou, L. Li, J. Liu, Y. Gao, Y. Yang, H. T. Shen, Distilling grounding dino for an edge-cloud collaborative advanced driver assistance system, IEEE Transactions on Circuits and Systems for Video Technology (2025) 1–1. doi:10.1109/TCSVT.2025.3586704. 31

2025

Pith tools

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

[1] [1]

C. Gao, W. Lei, X. He, M. de Rijke, T.-S. Chua, Advances and challenges in conversational recommender systems: A survey, AI Open 2 (2021) 100–126

work page 2021

[2] [2]

J. Zou, E. Kanoulas, P. Ren, Z. Ren, A. Sun, C. Long, Improving conversational recommender systems via transformer-based sequential modelling, in: Proceedings of the 45th International ACM SIGIR Con- ference on Research and Development in Information Retrieval, 2022, pp. 2319–2324

work page 2022

[3] [3]

Q. Chen, J. Lin, Y. Zhang, M. Ding, Y. Cen, H. Yang, J. Tang, To- wards knowledge-based recommender dialog system, in: Proceedings of 23 the 2019 Conference on Empirical Methods in Natural Language Pro- cessing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), 2019, pp. 1803–1813

work page 2019

[4] [4]

K. Zhou, W. X. Zhao, S. Bian, Y. Zhou, J.-R. Wen, J. Yu, Improving conversational recommender systems via knowledge graph based seman- tic fusion, in: Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2020, pp. 1006– 1014

work page 2020

[5] [5]

Sarkar, K

R. Sarkar, K. Goswami, M. Arcan, J. P. McCrae, Suggest me a movie for tonight: Leveraging knowledge graphs for conversational recommen- dation, in: Proceedings of the 28th International Conference on Com- putational Linguistics, 2020, pp. 4179–4189

work page 2020

[6] [6]

W. Ma, R. Takanobu, M. Huang, CR-walker: Tree-structured graph rea- soning and dialog acts for conversational recommendation, in: Proceed- ings of the 2021 Conference on Empirical Methods in Natural Language Processing, 2021, pp. 1839–1851

work page 2021

[7] [7]

K. Zhou, Y. Zhou, W. X. Zhao, X. Wang, J.-R. Wen, Towards topic- guided conversational recommender system, in: Proceedings of the 28th International Conference on Computational Linguistics, 2020, pp. 4128– 4139

work page 2020

[8] [8]

X. Wang, Y. Li, J. Liu, Conversation recommender system based on knowledge graph and time-series feature, in: The 5th International Conference on Computer Science and Application Engineering, 2021, pp. 1–5

work page 2021