Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

Multi-agents based User Values Mining for Recommendation

T0 review · 4 major / 4 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 The paper claims that stable user values mined by a zero-shot multi-LLM debate improve recommender accuracy, with contrastive fusion beating direct concatenation in every tested setting.

desk verdict A clever LLM-agent pipeline for mining user values, but the main recommendation result is fragile until the authors rule out test-item leakage and shore up the human evaluation. read the letter →

arxiv 2505.00981 v1 pith:JAFV6FQ3 submitted 2025-05-02 cs.IR

Lijian Chen , Wei Yuan , Tong Chen , Xiangyu Zhao , Nguyen Quoc Viet Hung , Hongzhi Yin This is my paper · ORCID
classification cs.IR
keywords uservaluesminingrecommendersystemslargelanguagemodelsmulti-agentdebatecontrastivelearningtextsummarizationzero-shotextractiontenuniversalbasic
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper sets out to show that a user's stable values, not just their transient clicks, can be recovered automatically from interaction history and put to work in recommender systems. It anchors values in the ten universal categories of a cross-cultural value taxonomy and extracts them with ZOOM, a zero-shot collaboration of LLM agents: evaluators summarize item content and generate candidate values from different orderings and decoding strategies, while supervisors review and debate the candidates until a consensus emerges. The paper claims this extraction aligns with human judgment far better than a naive LLM baseline, with F1 of 0.6813 on PENS and 0.7282 on MovieLens-1M, both more than six times the baseline, and that adding the mined values to two language-model recommenders improves Hit Rate and NDCG across all reported settings, with the contrastive-learning fusion outperforming simple concatenation. A sympathetic reader would care because it offers a path from noisy, short-term behavior signals to a stable representation of what a user actually values, which could make recommendations steadier and more explainable.

What carries the argument

The machinery is ZOOM's three-stage pipeline plus a contrastive fusion loss. First, an evaluator LLM summarizes each item's content so that more interactions fit into a limited context window, then generates candidate value sets from original-order and shuffled-order histories using four decoding strategies: beam search sampling, temperature sampling, top-k sampling, and nucleus sampling. The output space is a fixed set of ten universal basic values: achievement, power, hedonism, stimulation, self-direction, security, conformity, tradition, universalism, and benevolence. Second, supervisor LLMs built on different backbones, Llama-3-8B and Gemma-2-9B, review each evaluator's candidates and debate for up to three rounds, with a final supervisor resolving any non-consensus; the debate is designed to filter hallucinations and enforce consistency. Third, the extracted value labels are embedded and fused into the recommender either by direct concatenation or through a contrastive auxiliary loss that pulls the user preference embedding toward the user's own value embedding and away from the most similar unrelated users' value embeddings, with the loss weight and negative-sample count tuned per dataset.

What would settle it

Re-run the human evaluation with at least three independent annotators per user, a pre-registered coding manual for the ten value categories, and reported inter-annotator agreement: if agreement is low or ZOOM's F1 against the independent labels falls toward the baseline LLM's roughly 0.12, then the claim that ZOOM mines true user values is unsupported; a complementary control would replace the mined values with randomly assigned values in the same contrastive fusion and check whether the ranking gains persist, which would show the gains come from regularization rather than value content.

Watch

Extended reading notes

Core claim

The central claim is that user values, understood as stable long-term motivational goals rather than ephemeral interests, can be mined from historical item interactions by a zero-shot multi-agent LLM framework. The paper's evidence has two parts. First, ZOOM's extracted value sets match human annotations with F1 of 0.6813 on PENS and 0.7282 on MovieLens-1M, more than six times the 0.1160 and 0.1187 achieved by a base LLM, and ablations show that both text summarization and supervisor debate are required for that accuracy. Second, incorporating these values into the MoRec and EasyRec recommenders improves recommendation performance: direct concatenation helps in most settings, while the contrastive-learning fusion consistently beats both the original models and concatenation, with the largest gains on EasyRec with MovieLens-1M, up to 40.47 percent in HR@10 and 56.61 percent in NDCG@20. The paper interprets the larger gains for EasyRec as evidence that value embeddings align more naturally with text-based user representations than with sequence-based ones.

Load-bearing premise

The load-bearing premise is that a user's true values can be read off their click or rating history and that the human annotations used to score the extraction are reliable; the paper reports annotation on only 50 randomly selected users per dataset without stating how many annotators labelled each user, what instructions mapped interactions to the ten value categories, or how disagreements were resolved, so noisy ground truth would undermine the F1 claims and the whole value-mining premise.

Editorial extensions

If this is right

  • If the central claim holds, any text-grounded recommender can be augmented with a value module without retraining the agent pipeline, because ZOOM runs zero-shot before recommendation and only supplies side information.
  • Contrastive fusion is the safer integration choice: it improved results in every reported configuration, while direct concatenation occasionally dipped below baseline, for example HR@10 for EasyRec on PENS.
  • The larger gains on MovieLens-1M with EasyRec suggest value integration pays off most when the value distribution is diverse and user representations are already text-based.
  • The ablation results imply both summarization and debate are load-bearing: removing summarization drops F1 from 0.68 to 0.21 on PENS and from 0.72 to 0.32 on MovieLens-1M, and removing debate cuts F1 to less than half.
  • Deployment needs tuning of the negative-sample count to value diversity: skewed value distributions need fewer negatives, while diverse distributions need more.

Reading between the lines

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

  • A natural extension the paper does not test is whether the mined value vectors are stable over time: if values are genuinely stable, the same user's extracted values from two different time windows should match, which would validate the core premise independently of recommendation metrics.
  • The same evaluator-supervisor debate pipeline could be repurposed for other stable user attributes, such as personality traits or ethical preferences, whenever a fixed taxonomy exists.
  • If value embeddings carry the gains, a smaller distilled encoder might be trained to map interaction text directly to value embeddings, removing the cost of invoking several large LLMs per user at deployment.
  • The contrastive loss effectively regularizes user representations toward a psychologically meaningful semantic anchor, which suggests recommendations could become more robust to accidental or adversarial clicks, though the paper only implies this via its clickbait example.
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 / 4 minor

Summary. The paper proposes ZOOM, a zero-shot multi-LLM framework that mines Schwartz personal values from users' raw interaction histories. ZOOM condenses item text by summarization and then uses evaluator and supervisor agents, with a debate mechanism, to generate and refine value predictions. The mined values are integrated into two language-model-based recommender systems, MoRec and EasyRec, either by direct concatenation or by a contrastive-learning auxiliary loss. Experiments on PENS and MovieLens-1M report consistent improvements in HR@K and NDCG@K over the base recommenders, as well as F1 alignment scores between ZOOM's extracted values and human annotations.

Significance. If the results hold, the paper offers a practical and low-cost way to obtain stable user-value signals from textual interaction data, and the contrastive fusion strategy is simple enough to port to other recommender architectures. The framework is described in enough detail to be reproducible in principle, and the choice of two public datasets and two state-of-the-art baselines is appropriate. However, the central claims currently rest on two evaluation gaps: the train/test separation for the value-mining step is not specified, and the human annotation study that grounds the value-extraction accuracy is severely underspecified. These issues must be resolved before the reported gains can be attributed to user values rather than to leakage or to a generic auxiliary-text effect.

major comments (4)
  1. [Section 6.1.2 vs. Section 4.2 and Algorithm 1] The leave-one-out protocol in Section 6.1.2 reserves the most recent interaction as the test item, but the paper never states whether the raw interaction history H_u fed into Algorithm 1 excludes that held-out item. Section 4.2 and Algorithm 1 define ZOOM's input as the user's raw interaction history without any restriction. If H_u includes the test item, the test item's content enters the summarization, evaluator, and supervisor steps, and the value embedding used in Section 5 carries information about the very item whose rank is evaluated. This would make the Table 2 gains, such as EasyRec on MovieLens-1M improving HR@10 from 0.1957 to 0.2749, an artifact of test-label leakage rather than evidence for value-aware recommendation. The authors must explicitly state that value mining uses only the training-prefix interactions, and if that is not the case, the recommendation experiments must be rerun with the held-out item excluded from H_u.
  2. [Section 6.4 and Table 3] The claim that ZOOM's extracted values align with human judgment rests on an F1 evaluation of only 50 randomly selected users per dataset, yet the annotation protocol is not reported. The paper does not say how many annotators labeled each user, what instructions or definitions of the ten Schwartz values were provided, whether annotators saw the same interaction histories that ZOOM saw, how disagreements were resolved, or what the inter-annotator agreement was. Without this information, the F1 scores in Table 3 cannot be interpreted, and the premise that ZOOM mines true user values rather than plausible-sounding labels is unsupported. The authors should provide a complete annotation protocol and agreement statistics, or temper the grounding claim accordingly.
  3. [Section 5, Eq. (7), and Table 2] The experimental design does not isolate the effect of user-value semantics from the effect of any text-derived auxiliary signal. The contrastive loss aligns the user preference embedding with an embedding computed from the same interaction history used to train the recommender, so the gains in Table 2 could come from a regularization or self-supervision effect rather than from the specific content of Schwartz values. A control condition is needed, for example using randomly assigned value labels, generic text summaries unrelated to values, or value embeddings derived from a random subset of items, to show that the improvements are specifically attributable to the mined user values. As written, the causal claim 'incorporating user values ... consistently improves recommendation performance' is not established.
  4. [Section 6.3 and Table 2] All recommendation results in Table 2 are single point estimates with no error bars, multiple seeds, or significance tests. Several improvements are very small, such as MoRec on MovieLens-1M HR@20 (0.3792 to 0.3812, a 0.53% gain) and NDCG@20 (0.0455 to 0.0460, a 1.10% gain). Given the authors' claim that the improvements are consistent, the paper should report at least the variance across multiple random seeds or a standard significance test; otherwise, the claim of consistency across all scenarios is not statistically supported.
minor comments (4)
  1. [Section 5 and Section 6.2.3] The description of the direct concatenation method is internally inconsistent: Section 5 and the introduction say the user value vector is 'appended' or 'concatenated' to the user preference vector, while Section 6.2.3 says it is 'simply added' to the user preference embedding. Please clarify whether DC performs vector concatenation or vector addition, since this changes the implementation.
  2. [Section 6.5 and Table 3] The text describing the 'ZOOM w/o Sum' ablation says the F1 score drops to 0.36 on MovieLens-1M, but Table 3 reports 0.3200. The reported number in the text should be corrected to match the table.
  3. [Section 6.3] The sentence reporting improvements for the worst DC scenarios mixes HR@10 and NDCG@20 numbers in a way that is hard to parse: it claims 'improvements of 5.22% and 6.25% in HR@10, and 1.10% and 1.15% in NDCG@20' over Original and DC, but the mapping between the two scenarios and the four percentages is unclear. Please rephrase for clarity.
  4. [Equation (7)] The notation for the contrastive loss is imprecise: the set V, the embedding e_v^i, and the negative-set V^- are not formally defined before Eq. (7), and it is unclear whether the 'N embeddings with the lowest cosine similarity' are recomputed at every training step or fixed. Please define these objects and state the sampling frequency.

Circularity Check

1 steps flagged · score 6.0 of 10

ZOOM's value-mining input is not specified to exclude the leave-one-out test item, so the recommendation gains reduce partly to test-item leakage by the paper's own definitions.

  1. self definitional [Algorithm 1 (Input, line 1), Section 4.2 (Eq. 2 and text above it), Section 6.1.2 (Evaluation Protocol)]
    "Input: H_u: Raw interaction history of user u ... C_u←{LLM(P_sum, d_i)| d_i∈H_u} (Algorithm 1). 'for each user, the most recent interaction is reserved as the test item, while the remaining interactions are used for training' (Section 6.1.2)."

    ZOOM's final user-value vector v_u is computed from H_u via summarization and the evaluator/supervisor debate (Eqs. 2 and 6), and Algorithm 1 defines H_u as the user's raw interaction history with no stated exclusion of the held-out item. Section 6.1.2 makes the most recent interaction the test item, and that interaction is part of the raw history. The recommendation stage (Section 5) then enriches the user embedding with v_u via DC or CL and scores the held-out item. Therefore, unless H_u is explicitly restricted to the training prefix, the test item's content enters the model through the mined values, so the Table 2 improvements are not clean evidence for value-based recommendation but are partly attributable to test-item leakage.

full rationale

The rest of the derivation is self-contained: Schwartz's taxonomy is adopted from external work, ZOOM's summarization/debate machinery is not defined in terms of the recommendation objective, the human-annotation F1 study is an external (though small and protocol-under-specified) check, and the self-citations in the reference list ([4], [20], [56]) are peripheral and not load-bearing. The contrastive-learning fusion is a standard auxiliary objective rather than a renamed prediction. The one central circularity risk is the unstated overlap between H_u and the leave-one-out test item: Algorithm 1 defines H_u as the raw interaction history, while Section 6.1.2 defines the test item as the most recent interaction, which is part of that raw history. As written, the recommendation experiment therefore reduces partly to target-item leakage. If H_u is explicitly the training prefix, the paper would be essentially non-circular; as written, the central recommendation claim is compromised.

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

The paper is an empirical pipeline rather than a derivation. It rests on the Schwartz taxonomy as ground truth, on the assumption that LLM summaries preserve value-relevant semantics, on the belief that debate reduces hallucination, and on a small human annotation study. Hyperparameters lambda and N are tuned on validation, and the optimal N changes across datasets and models.

free parameters (4)
  • contrastive loss weight lambda = 0.01 (best of {0.0001, 0.001, 0.01})
    Tuned in RQ4 on validation; HR@10 is highest at lambda=0.01 for all settings.
  • number of negative samples N = Varies: PENS MoRec=3, PENS EasyRec=5, ML MoRec=7, ML EasyRec=9
    Tuned in RQ4 from {1,3,5,7,9}; the optimal value changes across datasets and base models.
  • contrastive temperature T = 0.1
    Fixed by hand in Section 6.2.3 without sensitivity analysis.
  • maximum debate rounds R = 3
    Fixed by hand in Section 6.2.2; no ablation on debate length is reported.
assumptions (4)
  • domain assumption Schwartz's Theory of Basic Values is a valid and complete taxonomy for representing user values in recommendation.
    Section 3.2 adopts the ten universal values and their conflict structure as the only value space for mining.
  • ad hoc to paper LLM text summarization preserves the value-relevant semantics of item content.
    Equation (2) compresses items so more history fits in the context window; no direct check that value signals survive compression is provided.
  • domain assumption Multi-LLM debate reduces hallucination and yields more accurate user values.
    Section 4.2.2 relies on the debate mechanism, citing prior work on multiagent debate; the assumption is not independently tested against other hallucination mitigations.
  • domain assumption Human annotators can reliably infer Schwartz values from short news or movie interaction histories.
    Section 6.4 uses human-annotated labels as ground truth for F1, but no annotation protocol, annotator count, or agreement measure is reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multi-agents based User Values Mining for Recommendation." pith.science (2026). https://pith.science/paper/JAFV6FQ3

@misc{pith2026250500981,
  author       = {Pith},
  title        = {Pith review of: Multi-agents based User Values Mining for Recommendation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JAFV6FQ3}},
  note         = {Machine review of arXiv:2505.00981}
}
read the original abstract

Recommender systems have rapidly evolved and become integral to many online services. However, existing systems sometimes produce unstable and unsatisfactory recommendations that fail to align with users' fundamental and long-term preferences. This is because they primarily focus on extracting shallow and short-term interests from user behavior data, which is inherently dynamic and challenging to model. Unlike these transient interests, user values are more stable and play a crucial role in shaping user behaviors, such as purchasing items and consuming content. Incorporating user values into recommender systems can help stabilize recommendation performance and ensure results better reflect users' latent preferences. However, acquiring user values is typically difficult and costly. To address this challenge, we leverage the strong language understanding, zero-shot inference, and generalization capabilities of Large Language Models (LLMs) to extract user values from users' historical interactions. Unfortunately, direct extraction using LLMs presents several challenges such as length constraints and hallucination. To overcome these issues, we propose ZOOM, a zero-shot multi-LLM collaborative framework for effective and accurate user value extraction. In ZOOM, we apply text summarization techniques to condense item content while preserving essential meaning. To mitigate hallucinations, ZOOM introduces two specialized agent roles: evaluators and supervisors, to collaboratively generate accurate user values. Extensive experiments on two widely used recommendation datasets with two state-of-the-art recommendation models demonstrate the effectiveness and generalization of our framework in automatic user value mining and recommendation performance improvement.

Figures

Figures reproduced from arXiv: 2505.00981 by the authors.

Figure 1
Figure 1. ZOOM Overview 2 RELATED WORK 2.1 Recommender Systems Over the past decades, recommender systems have significantly evolved and now play a crucial role in providing personalized recommendations. Early recommender systems primarily employed collaborative filtering (CF) [39], relying on unique identifiers (IDs) to represent users and items and predicting user preferences based on historical user-item interactions. A wi… view at source ↗
Figure 2
Figure 2. Overview of Schwartz’s Theory of Basic Values [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Text Summarization Prompt: P𝑠𝑢𝑚 the supervisors collaborate to reach a consensus on the final set of user values. In the following sections, we provide further details on the roles and processes of these two types of agents. 4.2.1 Evaluators. Since user value mining depends on users’ historically interacted items, which often contain lengthy content, it is inevitable that only a limited number of items can be proces… view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Evaluator Profile Prompt: P𝑒𝑣𝑎𝑙 evaluator to generate a diverse set of user values for each user through various reasoning paths (i.e., decoding strategies). We use S to indicate a set of decoding strategies employed to generate a diverse set of user values through var…
Figure 5
Figure 5. Figure 5: Supervisor Review Prompt: P𝑟𝑒𝑣𝑖𝑒𝑤 [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Supervisor Debate Prompt: P𝑑𝑒𝑏𝑎𝑡𝑒 outputs of this step are stored in their respective memory states, M𝑜𝑟𝑖 𝑠𝑢𝑝 [𝑚0] and M𝑟𝑎𝑛𝑑 𝑠𝑢𝑝 [𝑚0]. M𝑘 𝑠𝑢𝑝 [𝑚0] := 𝐿𝐿𝑀𝑘 𝑠𝑢𝑝 (P𝑟𝑒𝑣𝑖𝑒𝑤, E 𝑘 𝑢 ) , where 𝑘 ∈ {𝑜𝑟𝑖, 𝑟𝑎𝑛𝑑} (4) Then, the debate formally begins after this initial review (𝑟 = …
Figure 7
Figure 7. Figure 7: Supervisor Final Determination Prompt: P𝑓 𝑖𝑛𝑎𝑙 𝑟 (1 ≤ 𝑟 ≤ 𝑅), supervisors generate updated arguments by considering the debate prompt P𝑑𝑒𝑏𝑎𝑡𝑒 (as shown in [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 8
Figure 8. Figure 8: Analysis of Hyperparameters 𝜆 and 𝑁 with respect to the HR@10 is constrained to include only a few items (e.g., two or three) due to input length limits, resulting in a distorted view of the user’s interaction history and poor value extraction. Similarly, disabling the…
Figure 9
Figure 9. Figure 9: Word Cloud of User Values Extracted by ZOOM Based on Schwartz’s Theory of Basic Values. Top: [PITH_FULL_IMAGE:figures/full_fig_p020_9.png]
Figure 10
Figure 10. Figure 10: Case Study: Effectiveness of User Values Mining between a General LLM and ZOOM [PITH_FULL_IMAGE:figures/full_fig_p021_10.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. Autonomous Information Seeking: A Roadmap for Agentic Recommender Systems

    cs.IR 2026-07 accept novelty 5.0 of 10

    Agentic recommender systems are organized by agent role (assisted, as-recommender, as-simulator) crossed with autonomy levels L2–L5, yielding a roadmap of architectures, evaluation limits, and open challenges.

Reference graph

Works this paper leans on

71 extracted references · 18 canonical work pages · cited by 1 Pith paper

  1. [1]

    Xiang Ao, Xiting Wang, Ling Luo, Ying Qiao, Qing He, and Xing Xie. 2021. PENS: A dataset and generic framework for personalized news headline generation. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) . 82–92

    work page 2021

  2. [2]

    Mark J Brandt. 2017. Predicting ideological prejudice. Psychological Science 28, 6 (2017), 713–722

    work page 2017

  3. [3]

    Collin Burns, Haotian Ye, Dan Klein, and Jacob Steinhardt. 2022. Discovering latent knowledge in language models without supervision. arXiv preprint arXiv:2212.03827 (2022)

    arXiv 2022

  4. [4]

    Lijian Chen, Wei Yuan, Tong Chen, Guanhua Ye, Nguyen Quoc Viet Hung, and Hongzhi Yin. 2024. Adversarial item promotion on visually-aware recommender systems by guided diffusion. ACM Transactions on Information Systems 42, 6 (2024), 1–26

    work page 2024

  5. [5]

    Aakanksha Chowdhery, Sharan Narang, Jacob Devlin, Maarten Bosma, Gaurav Mishra, Adam Roberts, Paul Barham, Hyung Won Chung, Charles Sutton, Sebastian Gehrmann, et al. 2023. Palm: Scaling language modeling with pathways. Journal of Machine Learning Research 24, 240 (2023), 1–113

    2023

  6. [6]

    Paul F Christiano, Jan Leike, Tom Brown, Miljan Martic, Shane Legg, and Dario Amodei. 2017. Deep reinforcement learning from human preferences. Advances in neural information processing systems 30 (2017)

    2017

  7. [7]

    Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Yunxuan Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, et al. 2024. Scaling instruction-finetuned language models. Journal of Machine Learning Research 25, 70 (2024), 1–53

    2024

  8. [8]

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, et al. 2020. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929 (2020)

    arXiv 2020

Show all 71 references
  1. [9]

    Yilun Du, Shuang Li, Antonio Torralba, Joshua B Tenenbaum, and Igor Mordatch. 2023. Improving factuality and reasoning in language models through multiagent debate. In Forty-first International Conference on Machine Learning

    2023

  2. [10]

    Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, et al. 2024. The llama 3 herd of models. arXiv preprint arXiv:2407.21783 (2024)

    2024 arXiv

  3. [11]

    Angela Fan, Mike Lewis, and Yann Dauphin. 2018. Hierarchical neural story generation.arXiv preprint arXiv:1805.04833 (2018)

    2018 arXiv

  4. [12]

    Jessica Ficler and Yoav Goldberg. 2017. Controlling linguistic style aspects in neural language generation. arXiv preprint arXiv:1707.02633 (2017). ACM Trans. Inf. Syst., Vol. 1, No. 1, Article . Publication date: May 2024. Multi-agents based User Values Mining for Recommendation 23

    2017 arXiv

  5. [13]

    F Maxwell Harper and Joseph A Konstan. 2015. The movielens datasets: History and context. Acm transactions on interactive intelligent systems (tiis) 5, 4 (2015), 1–19

    2015

  6. [14]

    Ruining He and Julian McAuley. 2016. VBPR: visual bayesian personalized ranking from implicit feedback. InProceedings of the AAAI conference on artificial intelligence , Vol. 30

    2016

  7. [15]

    Xiangnan He, Lizi Liao, Hanwang Zhang, Liqiang Nie, Xia Hu, and Tat-Seng Chua. 2017. Neural collaborative filtering. In Proceedings of the 26th international conference on world wide web . 173–182

    2017

  8. [16]

    Balázs Hidasi, Alexandros Karatzoglou, Linas Baltrunas, and Domonkos Tikk. 2015. Session-based recommendations with recurrent neural networks. arXiv preprint arXiv:1511.06939 (2015)

    2015 arXiv

  9. [17]

    Ari Holtzman, Jan Buys, Li Du, Maxwell Forbes, and Yejin Choi. 2019. The curious case of neural text degeneration. arXiv preprint arXiv:1904.09751 (2019)

    2019 arXiv

  10. [18]

    Ari Holtzman, Jan Buys, Maxwell Forbes, Antoine Bosselut, David Golub, and Yejin Choi. 2018. Learning to write with cooperative discriminators. arXiv preprint arXiv:1805.06087 (2018)

    2018 arXiv

  11. [19]

    Lei Huang, Weijiang Yu, Weitao Ma, Weihong Zhong, Zhangyin Feng, Haotian Wang, Qianglong Chen, Weihua Peng, Xiaocheng Feng, Bing Qin, et al. 2023. A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions. ACM Transactions on Info...

    2023

  12. [20]

    Nguyen Quoc Viet Hung, Huynh Huu Viet, Nguyen Thanh Tam, Matthias Weidlich, Hongzhi Yin, and Xiaofang Zhou

  13. [21]

    Gabriel Jaskolka, Janice M Beyer, and Harrison M Trice. 1985. Measuring and predicting managerial success. Journal of vocational behavior 26, 2 (1985), 189–205

    1985

  14. [22]

    Ziwei Ji, Nayeon Lee, Rita Frieske, Tiezheng Yu, Dan Su, Yan Xu, Etsuko Ishii, Ye Jin Bang, Andrea Madotto, and Pascale Fung. 2023. Survey of hallucination in natural language generation. Comput. Surveys 55, 12 (2023), 1–38

    2023

  15. [23]

    Wang-Cheng Kang and Julian McAuley. 2018. Self-attentive sequential recommendation. In 2018 IEEE international conference on data mining (ICDM) . IEEE, 197–206

    2018

  16. [24]

    Yehuda Koren, Robert Bell, and Chris Volinsky. 2009. Matrix factorization techniques for recommender systems. Computer 42, 8 (2009), 30–37

    2009

  17. [25]

    Walid Krichene and Steffen Rendle. 2020. On sampled metrics for item recommendation. In Proceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining . 1748–1757

    2020

  18. [26]

    Gonzalez, Hao Zhang, and Ion Stoica

    Woosuk Kwon, Zhuohan Li, Siyuan Zhuang, Ying Sheng, Lianmin Zheng, Cody Hao Yu, Joseph E. Gonzalez, Hao Zhang, and Ion Stoica. 2023. Efficient Memory Management for Large Language Model Serving with PagedAttention. In Proceedings of the ACM SIGOPS 29th Symposium on Operating S...

    2023

  19. [27]

    Qingyao Li, Lingyue Fu, Weiming Zhang, Xianyu Chen, Jingwei Yu, Wei Xia, Weinan Zhang, Ruiming Tang, and Yong Yu. 2023. Adapting large language models for education: Foundational capabilities, potentials, and challenges. arXiv preprint arXiv:2401.08664 (2023)

    2023 arXiv

  20. [28]

    Youhua Li, Hanwen Du, Yongxin Ni, Pengpeng Zhao, Qi Guo, Fajie Yuan, and Xiaofang Zhou. 2024. Multi-modality is all you need for transferable recommender systems. In 2024 IEEE 40th International Conference on Data Engineering (ICDE). IEEE, 5008–5021

    2024

  21. [29]

    Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692 (2019)

    2019 arXiv

  22. [30]

    Måns Magnusson, Michael Andersen, Johan Jonasson, and Aki Vehtari. 2019. Bayesian leave-one-out cross-validation for large data. In International Conference on Machine Learning . PMLR, 4244–4253

    2019

  23. [31]

    Ben Mann, N Ryder, M Subbiah, J Kaplan, P Dhariwal, A Neelakantan, P Shyam, G Sastry, A Askell, S Agarwal, et al

  24. [32]

    Joshua Maynez, Shashi Narayan, Bernd Bohnet, and Ryan McDonald. 2020. On faithfulness and factuality in abstractive summarization. arXiv preprint arXiv:2005.00661 (2020)

    2020 arXiv

  25. [33]

    Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. 2022. Training language models to follow instructions with human feedback. Advances in neural information processing systems 35...

    2022

  26. [34]

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. 2019. Pytorch: An imperative style, high-performance deep learning library. Advances in neural information processing system...

    2019

  27. [35]

    Baolin Peng, Chunyuan Li, Pengcheng He, Michel Galley, and Jianfeng Gao. 2023. Instruction tuning with gpt-4. arXiv preprint arXiv:2304.03277 (2023)

    2023 arXiv

  28. [36]

    Xubin Ren and Chao Huang. 2024. EasyRec: Simple yet Effective Language Models for Recommendation.arXiv preprint arXiv:2408.08821 (2024). ACM Trans. Inf. Syst., Vol. 1, No. 1, Article . Publication date: May 2024. 24 Chen et al

    2024

  29. [37]

    Steffen Rendle, Christoph Freudenthaler, Zeno Gantner, and Lars Schmidt-Thieme. 2012. BPR: Bayesian personalized ranking from implicit feedback. arXiv preprint arXiv:1205.2618 (2012)

    2012 arXiv

  30. [38]

    Steffen Rendle, Christoph Freudenthaler, and Lars Schmidt-Thieme. 2010. Factorizing personalized markov chains for next-basket recommendation. In Proceedings of the 19th international conference on World wide web . 811–820

    2010

  31. [39]

    Badrul Sarwar, George Karypis, Joseph Konstan, and John Riedl. 2001. Item-based collaborative filtering recommenda- tion algorithms. In Proceedings of the 10th international conference on World Wide Web . 285–295

    2001

  32. [40]

    J Ben Schafer, Joseph Konstan, and John Riedl. 1999. Recommender systems in e-commerce. In Proceedings of the 1st ACM conference on Electronic commerce . 158–166

    1999

  33. [41]

    Shalom H Schwartz. 2012. An overview of the Schwartz theory of basic values. Online readings in Psychology and Culture 2, 1 (2012), 11

    2012

  34. [42]

    Nisan Stiennon, Long Ouyang, Jeffrey Wu, Daniel Ziegler, Ryan Lowe, Chelsea Voss, Alec Radford, Dario Amodei, and Paul F Christiano. 2020. Learning to summarize with human feedback. Advances in neural information processing systems 33 (2020), 3008–3021

    2020

  35. [43]

    Ruixiang Tang, Xiaotian Han, Xiaoqian Jiang, and Xia Hu. 2023. Does synthetic data generation of llms help clinical text mining? arXiv preprint arXiv:2303.04360 (2023)

    2023 arXiv

  36. [44]

    Yi Tay, Luu Anh Tuan, and Siu Cheung Hui. 2018. Latent relational metric learning via memory-based attention for collaborative ranking. In Proceedings of the 2018 world wide web conference . 729–739

    2018

  37. [45]

    Gemma Team, Morgane Riviere, Shreya Pathak, Pier Giuseppe Sessa, Cassidy Hardin, Surya Bhupatiraju, Léonard Hussenot, Thomas Mesnard, Bobak Shahriari, Alexandre Ramé, et al. 2024. Gemma 2: Improving open language models at a practical size. arXiv preprint arXiv:2408.00118 (2024)

    2024 arXiv

  38. [46]

    Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, et al. 2023. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288 (2023)

    2023 arXiv

  39. [47]

    Johannes Treutlein, Dami Choi, Jan Betley, Samuel Marks, Cem Anil, Roger B Grosse, and Owain Evans. 2024. Connecting the dots: Llms can infer and verbalize latent structure from disparate training data. Advances in Neural Information Processing Systems 37 (2024), 140667–140730

    2024

  40. [48]

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. Advances in neural information processing systems 30 (2017)

    2017

  41. [49]

    Ashwin K Vijayakumar, Michael Cogswell, Ramprasath R Selvaraju, Qing Sun, Stefan Lee, David Crandall, and Dhruv Batra. 2016. Diverse beam search: Decoding diverse solutions from neural sequence models. arXiv preprint arXiv:1610.02424 (2016)

    2016 arXiv

  42. [50]

    Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc Le, Ed Chi, Sharan Narang, Aakanksha Chowdhery, and Denny Zhou. 2022. Self-consistency improves chain of thought reasoning in language models. arXiv preprint arXiv:2203.11171 (2022)

    2022 arXiv

  43. [51]

    Xinyuan Wang, Liang Wu, Liangjie Hong, Hao Liu, and Yanjie Fu. 2024. Llm-enhanced user-item interactions: Leveraging edge information for optimized recommendations. arXiv preprint arXiv:2402.09617 (2024)

    2024 arXiv

  44. [52]

    Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, et al . 2022. Chain-of-thought prompting elicits reasoning in large language models. Advances in neural information processing systems 35 (2022), 24824–24837

    2022

  45. [53]

    Wei Wei, Xubin Ren, Jiabin Tang, Qinyong Wang, Lixin Su, Suqi Cheng, Junfeng Wang, Dawei Yin, and Chao Huang

  46. [54]

    Shijie Wu, Ozan Irsoy, Steven Lu, Vadim Dabravolski, Mark Dredze, Sebastian Gehrmann, Prabhanjan Kambadur, David Rosenberg, and Gideon Mann. 2023. Bloomberggpt: A large language model for finance.arXiv preprint arXiv:2303.17564 (2023)

    2023 arXiv

  47. [55]

    Jing Yao, Xiaoyuan Yi, Xiting Wang, Yifan Gong, and Xing Xie. 2023. Value fulcra: Mapping large language models to the multidimensional spectrum of basic human values. arXiv preprint arXiv:2311.10766 (2023)

    2023 arXiv

  48. [56]

    Junliang Yu, Xin Xia, Tong Chen, Lizhen Cui, Nguyen Quoc Viet Hung, and Hongzhi Yin. 2023. XSimGCL: Towards extremely simple graph contrastive learning for recommendation.IEEE Transactions on Knowledge and Data Engineering 36, 2 (2023), 913–926

    2023

  49. [57]

    Junliang Yu, Hongzhi Yin, Xin Xia, Tong Chen, Lizhen Cui, and Quoc Viet Hung Nguyen. 2022. Are graph augmentations necessary? simple graph contrastive learning for recommendation. In Proceedings of the 45th international ACM SIGIR conference on research and development in info...

    2022

  50. [58]

    Junliang Yu, Hongzhi Yin, Xin Xia, Tong Chen, Jundong Li, and Zi Huang. 2023. Self-supervised learning for recommender systems: A survey. IEEE Transactions on Knowledge and Data Engineering 36, 1 (2023), 335–355

    2023

  51. [59]

    Wei Yuan, Chaoqun Yang, Liang Qu, Nguyen Quoc Viet Hung, Guanhua Ye, and Hongzhi Yin. 2024. PTF-FSR: A Parameter Transmission-Free Federated Sequential Recommender System. ACM Transactions on Information Systems ACM Trans. Inf. Syst., Vol. 1, No. 1, Article . Publication date:...

    2024

  52. [60]

    Wei Yuan, Chaoqun Yang, Guanhua Ye, Tong Chen, Nguyen Quoc Viet Hung, and Hongzhi Yin. 2024. FELLAS: Enhancing Federated Sequential Recommendation with LLM as External Services. ACM Transactions on Information Systems (2024)

    2024

  53. [61]

    Zheng Yuan, Fajie Yuan, Yu Song, Youhua Li, Junchen Fu, Fei Yang, Yunzhu Pan, and Yongxin Ni. 2023. Where to go next for recommender systems? id-vs. modality-based recommender models revisited. In Proceedings of the 46th International ACM SIGIR Conference on Research and Devel...

    2023

  54. [62]

    An Zhang, Yuxin Chen, Leheng Sheng, Xiang Wang, and Tat-Seng Chua. 2024. On generative agents in recommendation. In Proceedings of the 47th international ACM SIGIR conference on research and development in Information Retrieval . 1807–1817

    2024

  55. [63]

    Shengyu Zhang, Linfeng Dong, Xiaoya Li, Sen Zhang, Xiaofei Sun, Shuhe Wang, Jiwei Li, Runyi Hu, Tianwei Zhang, Fei Wu, et al. 2023. Instruction tuning for large language models: A survey. arXiv preprint arXiv:2308.10792 (2023)

    2023

  56. [64]

    Tianyi Zhang, Faisal Ladhak, Esin Durmus, Percy Liang, Kathleen McKeown, and Tatsunori B Hashimoto. 2024. Benchmarking large language models for news summarization. Transactions of the Association for Computational Linguistics 12 (2024), 39–57

    2024

  57. [65]

    Wayne Xin Zhao, Zihan Lin, Zhichao Feng, Pengfei Wang, and Ji-Rong Wen. 2022. A revisiting study of appropriate offline evaluation for top-N recommendation algorithms. ACM Transactions on Information Systems 41, 2 (2022), 1–41

    2022

  58. [66]

    Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zi Lin, Zhuohan Li, Dacheng Li, Eric Xing, et al. 2023. Judging llm-as-a-judge with mt-bench and chatbot arena. Advances in Neural Information Processing Systems 36 (2023), 46595–46623

    2023

  59. [67]

    Chunting Zhou, Pengfei Liu, Puxin Xu, Srinivasan Iyer, Jiao Sun, Yuning Mao, Xuezhe Ma, Avia Efrat, Ping Yu, Lili Yu, et al. 2023. Lima: Less is more for alignment. Advances in Neural Information Processing Systems 36 (2023), 55006–55021

    2023

  60. [68]

    Yutao Zhu, Huaying Yuan, Shuting Wang, Jiongnan Liu, Wenhan Liu, Chenlong Deng, Haonan Chen, Zheng Liu, Zhicheng Dou, and Ji-Rong Wen. 2023. Large language models for information retrieval: A survey. arXiv preprint arXiv:2308.07107 (2023). ACM Trans. Inf. Syst., Vol. 1, No. 1,...

    2023

  61. [2017]

    IEEE Transactions on Knowledge and Data Engineering 30, 1 (2017), 1–14

    Computing crowd consensus with partial agreement. IEEE Transactions on Knowledge and Data Engineering 30, 1 (2017), 1–14

    2017

  62. [2020]

    arXiv preprint arXiv:2005.14165 1 (2020), 3

    Language models are few-shot learners. arXiv preprint arXiv:2005.14165 1 (2020), 3

    2020 arXiv

  63. [2024]

    In Proceedings of the 17th ACM International Conference on Web Search and Data Mining

    Llmrec: Large language models with graph augmentation for recommendation. In Proceedings of the 17th ACM International Conference on Web Search and Data Mining . 806–815

Pith tools

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