Pith. sign in

REVIEW 2 major objections 5 minor 48 references

Prompt Tuning for Item Cold-start Recommendation

T0 review · 2 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Positive feedback beats text prompts for cold-start items

desk verdict Useful industrial prompt-tuning idea, but the offline evaluation as written leaks the test user's ID into the item prompt; the core SOTA claim is not established without a temporal exclusion. read the letter →

arxiv 2412.18082 v1 pith:EKAV4MAN submitted 2024-12-24 cs.IR cs.AI

Yuezihan Jiang , Gaode Chen , Wenhan Zhang , Jingchi Wang , Yinjie Jiang , Qi Zhang , Jingjian Lin , Peng Jiang
show 1 more author
Kaigui Bian
This is my paper · ORCID
classification cs.IRcs.AI
keywords itemcold-startrecommendationprompttuningpinnaclefeedbackpersonalizednetworkCTRpredictionmodelbiasprompt-enhancedlosssequential
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

Prompt-tuning for cold-start recommendations usually feeds the model text descriptions or content features, which are costly to annotate and semantically remote from the click-prediction task. PROMO instead builds each new item's prompt from its high-value positive feedback, meaning the users who most strongly engaged with the item, and encodes that feedback through per-item personalized prompt networks while freezing the pretrained backbone. The paper argues this removes both problems at once: feedback is directly task-relevant, and per-item networks stop popular items from dominating the tuning updates. On four public datasets it reports consistent gains over prior cold-start and prompt-based methods, and a live A/B test on a large short-video platform reports higher click rate, play time, likes, and collections for cold-start items. The paper is trying to establish that positive feedback is the right prompt signal for item cold-start recommendation and that it can be deployed cheaply at scale.

What carries the argument

Two mechanisms carry the argument. First, pinnacle feedback as prompt data: for each item, the top-k users by a weighted score of dwell time and interaction such as like, follow, or forward form the positive list, and k non-interacting users form the negative list, with their pretrained ID embeddings serving as the prompt. Items without direct feedback borrow the most similar popular item's pinnacle list as pseudo-prompt information. Second, the personalized prompt network: a learnable prompt embedding is reshaped into the weights and biases of a small MLP for that item alone, so the prompt information is encoded by item-specific parameters rather than a shared network. The main losses are a pairwise gap between the encoded pinnacle and negative feedback representations plus a batch-level loss that pushes cold-start positive scores above warm-item negative scores, while the frozen pretrained backbone contributes the final item embedding through a fusion MLP.

What would settle it

Re-run the MovieLens and KuaiRand experiments with a strict temporal split: for each user, build each test item's pinnacle feedback list only from interactions timestamped before the user's last interaction, and check whether PROMO still beats the baselines. If the gains shrink to noise, the claimed advantage of positive feedback as prompt comes from label leakage rather than task-relevant information.

Watch

Extended reading notes

Core claim

The central claim is that, for item cold-start recommendation, the most informative prompt is not an item's content description but its pinnacle feedback: the small set of users who gave the strongest positive reactions to the item. Treating these users' ID embeddings as prompt input recasts user-item matching as user-user matching and supplies interest signals that content features lack. To prevent warm-up items from dominating the tuned parameters, PROMO generates a separate prompt network for each item from a learnable prompt embedding, so each item only updates its own network. Two auxiliary losses, the pinnacle-feedback prompt-enhanced loss and the intra-batch popularity-aware prompt-enhanced loss, sharpen the separation between cold-start positives and negatives and between cold-start positives and popular-item negatives. The paper reports that this design outperforms prior content-prompt, collaborative-filtering, meta-learning, and prompt baselines on MovieLens, KuaiRand, and TMall, and that the deployed system lifted cold-start click rate by 3.2%, play time by 4.8%, likes by 3.9%, and collections by 4.0% in a 14-day A/B test.

Load-bearing premise

The offline evaluation assumes that the pinnacle feedback list used to build a test item's embedding is constructed without including the held-out test user's own interaction with that item; otherwise the prediction is partly computed from the label it is supposed to predict.

Editorial extensions

If this is right

  • If positive feedback is the right prompt signal, cold-start recommendation can be improved without manual annotation or text descriptions for new items.
  • Per-item prompt networks make the tuning phase parameter-efficient: the paper reports tuning only 17.7% to 27.6% of the parameters used by full fine-tuning.
  • The pinnacle-feedback formulation reframes item cold-start as user-user matching, which could help cold-start items gain traffic earlier and transition to popular status.
  • The popularity-aware loss provides a concrete recipe for counteracting warm-item bias in two-tower CTR models.
  • The reported A/B gains suggest the method can move from offline benchmarks to a billion-user production recommender.

Reading between the lines

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

  • Extension: the same pinnacle-feedback prompt idea could be applied to user cold-start by swapping roles, using high-affinity items as the prompt for a new user.
  • Extension: a strict temporal split of feedback lists would test whether the offline gains survive when the held-out user's own engagement is excluded from the prompt; this is the main threat the current leave-one-out setup leaves open.
  • Extension: the dwell-time-and-interaction scoring rule for selecting pinnacle feedback could be replaced with a learned scorer, which may help in e-commerce domains where explicit signals differ.
  • Connection: representing an item as a function of its best users links PROMO to neighbor-based collaborative filtering ideas, suggesting it could be combined with item-item similarity graphs.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 5 minor

Summary. PROMO is a prompt-tuning method for item cold-start CTR prediction. It constructs a 'pinnacle feedback' prompt as the top-k users who gave positive feedback on an item, encodes these user IDs through item-wise personalized prompt networks, and combines the result with base-model embeddings to obtain the final item representation. The paper claims improvements over state-of-the-art methods on MovieLens, KuaiRand, and TMall, and reports a large-scale online A/B test at Kuaishou. The core idea is parameter-efficient and avoids manual annotation. However, the offline evaluation suffers from label leakage: the pinnacle feedback list for a test item can contain the held-out user, so the test label is used in computing the prediction score. This undermines the central empirical claims.

Significance. The manuscript addresses a practically important problem, and the proposed mechanism is original in using user IDs of positive feedback as prompts. It also provides a deployed system and code. If the evaluation were clean, the SOTA claims would be significant. Because the offline results are confounded by direct label leakage and the online results lack statistical detail, the significance is not established in the current version.

major comments (2)
  1. [Sec. 4.1.1, Sec. 5.1, Eq. (10)] The construction of the pinnacle feedback list Pos_i in Sec. 4.1.1 is defined for all 'users who have viewed the item i' with no temporal or held-out restriction. Sec. 5.1 uses leave-one-out, taking each user's last interaction as the test item. Therefore, for a test pair (u,i), user u is a viewer of i and is in general a candidate for Pos_i; under Eq. (10), the term (1/k) * sum(S_pos_i) then contains the test user's own ID embedding e_u. The final score <e_final_u, e_final_i> is thus inflated by an inner product with e_u itself, i.e., the model's prediction uses the ground-truth label. This is direct label leakage and invalidates the headline results in Tables 2-4 and the claim in the abstract that positive feedback is more suitable as prompt information. The paper must specify and implement a temporal split that excludes the test interaction from Pos_i (and Neg_i), and re-report all results; the current text does not rule out the leak.
  2. [Tables 2 and 5] No confidence intervals, standard errors, or significance tests are reported for any of the offline metrics in Table 2, even though PROMO's gains over SASRec on some datasets (e.g., MovieLens 1M) are small (57.5 vs 56.9 H@5). The online A/B test in Table 5 reports point estimates (+3.2% to +4.8%) without confidence intervals or p-values, and the baseline is described only as 'similar to the SASRec baseline method.' These omissions make it impossible to assess whether the claimed improvements are statistically reliable; the paper should provide bootstrap CIs or significance tests for the offline comparisons and a proper online significance analysis.
minor comments (5)
  1. [Abstract] The abstract contains the typo 'pinnaclce feedback'.
  2. [Sec. 5.3] The text states that results are reported for H@1, H@5, H@10, N@1, N@5, and N@10, but Table 2 only contains H@5, H@10, N@5, and N@10; the claimed H@1 and N@1 are missing.
  3. [Sec. 2.2] The baseline descriptions for SASRec and DSSM are identical; the DSSM bullet appears to be a copy of the SASRec bullet and should be corrected.
  4. [Sec. 4.1.1] Eq. (4) uses CR_{u,i} and IR_{u,i} but these quantities are not formally defined, and it is unclear how they are computed for MovieLens and TMall, which lack staying time and interaction scores.
  5. [Sec. 5.8] The online deployment section does not state the exact size of each group beyond 'more than 30 million users' and does not precisely identify the baseline serving model, which limits reproducibility.

Circularity Check

1 steps flagged · score 8.0 of 10

The offline 'prediction' uses the test user's own ID: Pos_i is built from all viewers with no temporal split, then Eq. (10) averages Pos_i into the item embedding.

  1. self definitional [Sec. 4.1.1 (Prompt Data), Eq. (10), Sec. 5.1 (Datasets)]
    "“For each user u ∈ U who has viewed the item i, we consider multiple positive feedback ... we select the top-k users with the highest value as the pinnacle sample list for the item i, that is, Pos_i = (u_pos_1, ..., u_pos_k).” ... “Following the same leave-one-out technique, we take the last interactive item for each user as the test data.” ... “e_final_i = MLP(h_i, 1/k Σ S_pos_i, e_p_i)” and “y_hat_u,i = [e_final_u]^T · e_final_i.”"

    The test user is a viewer of the held-out item by construction of the leave-one-out split, and no temporal split is stated for Pos_i. Eq. (10) puts the average of the pinnacle users' ID embeddings into e_final_i, so for the test pair (u,i), e_final_i contains e_u. The final score is then the inner product of e_final_u with a vector containing e_u, meaning the prediction is partially computed from the ground-truth user-item interaction itself. The comparison to PROMO-I/F/IF is also forced: those variants do not include the test user's ID in the item prompt, so PROMO's large gains are the signature of self-matching rather than of positive feedback semantics.

full rationale

The central offline claim—that pinnacle positive feedback is more suitable than content descriptions and yields SOTA cold-start performance—rests on an evaluation construction that is circular. Pos_i is defined as the top-k users among all users who viewed the item, with no exclusion of the held-out test user. Under the leave-one-out protocol, every test user is a viewer of their test item, so the test user's own ID enters the item prompt average in Eq. (10) by definition. The final score is then an inner product with a vector containing the test user's embedding, so the 'predicted' interaction is partly computed from the ground-truth label itself. This explains the pattern in Table 3: replacing user-ID pinnacle feedback with item IDs or features removes the self-matching channel and drops performance sharply. The paper does not state that Pos_i is built with a temporal split or with test users removed, so the text as written exhibits a direct reduction of the prediction to its input. I found no separate load-bearing self-citation chain or imported uniqueness theorem; the circularity is in the offline evaluation, not in the related-work citations. Because the primary empirical proof and the main SOTA claims reduce by construction, the circularity score is 8.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The method depends on several hand-set hyperparameters and on assumptions about the informativeness of positive-feedback user IDs and the reliability of item similarity for pseudo-pinnacle construction. The most consequential unstated premise is that the evaluation split excludes the held-out user from the prompt; this is not written in the paper.

free parameters (6)
  • alpha = not reported
    Weight on staying time CR in Eq. (4) for pinnacle feedback selection; the paper gives no value or sensitivity analysis.
  • beta = not reported
    Weight on interaction score IR in Eq. (4); not reported.
  • k = not reported
    Number of positive and negative feedback samples per item in Pos_i and Neg_i; no value or ablation reported.
  • lambda_1 = not reported
    Weight for pinnacle feedback prompt-enhanced loss in Eq. (13); no value reported.
  • lambda_2 = not reported
    Weight for popularity-aware prompt-enhanced loss in Eq. (13); no value reported.
  • cold-start popularity threshold = 20 / 50 / 50 / 20
    Per-dataset threshold separating cold-start from popular items; chosen by hand and used to define the evaluation set.
assumptions (5)
  • domain assumption The pretrained SASRec model provides valid user and item embeddings for the cold-start task.
    PROMO freezes this backbone and builds all prompts and pseudo-pinnacle similarities on its representations (Section 4).
  • domain assumption Positive feedback user IDs are task-relevant and encode user interest better than content features.
    The motivating experiment in Section 1 compares two dual-tower models and asserts this without a formal proof or error bars.
  • domain assumption Item similarity in Eq. (5) is a reliable basis for transferring feedback from popular to cold items.
    Pseudo-pinnacle samples for items without feedback are taken from the single most similar popular item; no validation of this transfer is given.
  • ad hoc to paper Pinnacle feedback prompts are constructed without the held-out test user.
    The paper does not state a temporal exclusion when building Pos_i in Section 4.1.1, yet the leave-one-out evaluation in Section 5.1 depends on it to avoid label leakage.
  • domain assumption Leave-one-out evaluation with random negative sampling approximates online cold-start performance.
    Standard in the literature, but the online setting differs in feedback timing and item age.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Prompt Tuning for Item Cold-start Recommendation." pith.science (2026). https://pith.science/paper/EKAV4MAN

@misc{pith2026241218082,
  author       = {Pith},
  title        = {Pith review of: Prompt Tuning for Item Cold-start Recommendation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EKAV4MAN}},
  note         = {Machine review of arXiv:2412.18082}
}
read the original abstract

The item cold-start problem is crucial for online recommender systems, as the success of the cold-start phase determines whether items can transition into popular ones. Prompt learning, a powerful technique used in natural language processing (NLP) to address zero- or few-shot problems, has been adapted for recommender systems to tackle similar challenges. However, existing methods typically rely on content-based properties or text descriptions for prompting, which we argue may be suboptimal for cold-start recommendations due to 1) semantic gaps with recommender tasks, 2) model bias caused by warm-up items contribute most of the positive feedback to the model, which is the core of the cold-start problem that hinders the recommender quality on cold-start items. We propose to leverage high-value positive feedback, termed pinnacle feedback as prompt information, to simultaneously resolve the above two problems. We experimentally prove that compared to the content description proposed in existing works, the positive feedback is more suitable to serve as prompt information by bridging the semantic gaps. Besides, we propose item-wise personalized prompt networks to encode pinnaclce feedback to relieve the model bias by the positive feedback dominance problem. Extensive experiments on four real-world datasets demonstrate the superiority of our model over state-of-the-art methods. Moreover, PROMO has been successfully deployed on a popular short-video sharing platform, a billion-user scale commercial short-video application, achieving remarkable performance gains across various commercial metrics within cold-start scenarios

Figures

Figures reproduced from arXiv: 2412.18082 by the authors.

Figure 1
Figure 1. The model accuracy on cold-start items with item [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 3
Figure 3. Overview of PROMO. PROMO utilizes item features and pinnacle feedback as the prompt information to generate prompt embedding and personalized prompt network, then PROMO optimizes the prompt embedding and eliminates the model bias by the pinnacle feedback prompt-enhance loss and the popularity-aware prompt-enhanced loss separately. original sparse features into low-dimensional vectors. For user 𝑢 ∈ U and item 𝑖 ∈ I, … view at source ↗
Figure 4
Figure 4. The visualization of item representation involves [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: The memory retention rate corresponds with the [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 26 canonical work pages

  1. [1]

    Talha Mahboob Alam and Mazhar Javed Awan. 2018. Domain analysis of infor- mation extraction techniques. Int. J. Multidiscip. Sci. Eng 9, 6 (2018)

    work page 2018

  2. [2]

    Talha Mahboob Alam, Kamran Shaukat, Mubbashar Mushtaq, Yasir Ali, Matloob Khushi, Suhuai Luo, and Abdul Wahab. 2021. Corporate bankruptcy prediction: An approach towards better corporate world.Comput. J. 64, 11 (2021), 1731–1746

    work page 2021

  3. [3]

    Oren Barkan, Noam Koenigstein, Eylon Yogev, and Ori Katz. 2019. CB2CF: a neu- ral multiview content-to-collaborative filtering model for completely cold item recommendations. In Proceedings of the 13th ACM Conference on Recommender Systems. 228–236

    work page 2019

  4. [4]

    Dmitry Bogdanov, Martín Haro, Ferdinand Fuhrmann, Anna Xambó, Emilia Gómez, and Perfecto Herrera. 2013. Semantic audio content-based music recom- mendation and visualization based on user preference examples. Information Processing & Management 49, 1 (2013), 13–33

    work page 2013

  5. [5]

    Yuwei Cao, Liangwei Yang, Chen Wang, Zhiwei Liu, Hao Peng, Chenyu You, and Philip S Yu. 2023. Multi-task item-attribute graph pre-training for strict cold-start item recommendation. In Proceedings of the 17th ACM Conference on Recommender Systems. 322–333

    work page 2023

  6. [6]

    Gaode Chen, Xinghua Zhang, Yijun Su, Yantong Lai, Ji Xiang, Junbo Zhang, and Yu Zheng. 2023. Win-win: a privacy-preserving federated framework for dual-target cross-domain recommendation. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 37. 4149–4156

    2023

  7. [7]

    Gaode Chen, Xinghua Zhang, Yanyan Zhao, Cong Xue, and Ji Xiang. 2021. Ex- ploring Periodicity and Interactivity in Multi-Interest Framework for Sequential Recommendation. In Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence. 1426–1433

    work page 2021

  8. [8]

    Ke Chen, Beici Liang, Xiaoshuan Ma, and Minwei Gu. 2021. Learning audio embeddings with user listening data for content-based music recommendation. In ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 3015–3019

    work page 2021

Show all 48 references
  1. [9]

    Yifan Chen, Devamanyu Hazarika, Mahdi Namazifar, Yang Liu, Di Jin, and Dilek Hakkani-Tur. 2022. Inducer-tuning: Connecting Prefix-tuning and Adapter- tuning. arXiv preprint arXiv:2210.14469 (2022)

    2022 arXiv

  2. [10]

    Yashar Deldjoo, Markus Schedl, Paolo Cremonesi, Gabirella Pasi, et al . 2018. Content-based multimedia recommendation systems: definition and application domains. In Italian Information Retrieval Workshop. 1–4

    2018

  3. [11]

    Ning Ding, Shengding Hu, Weilin Zhao, Yulin Chen, Zhiyuan Liu, Hai-Tao Zheng, and Maosong Sun. 2021. Openprompt: An open-source framework for prompt- learning. arXiv preprint arXiv:2111.01998 (2021)

    2021 arXiv

  4. [12]

    Chongming Gao, Shijun Li, Yuan Zhang, Jiawei Chen, Biao Li, Wenqiang Lei, Peng Jiang, and Xiangnan He. 2022. KuaiRand: An Unbiased Sequential Rec- ommendation Dataset with Randomly Exposed Videos. In Proceedings of the 31st ACM International Conference on Information & Knowled...

    2022

  5. [13]

    Shijie Geng, Shuchang Liu, Zuohui Fu, Yingqiang Ge, and Yongfeng Zhang. 2022. Recommendation as language processing (rlp): A unified pretrain, personalized prompt & predict paradigm (p5). In Proceedings of the 16th ACM Conference on Recommender Systems. 299–315

    2022

  6. [14]

    Huifeng Guo, Ruiming Tang, Yunming Ye, Zhenguo Li, and Xiuqiang He. 2017. DeepFM: a factorization-machine based neural network for CTR prediction.arXiv preprint arXiv:1703.04247 (2017)

    2017 arXiv

  7. [15]

    Bowen Hao, Hongzhi Yin, Jing Zhang, Cuiping Li, and Hong Chen. 2023. A multi-strategy-based pre-training method for cold-start recommendation. ACM Transactions on Information Systems 41, 2 (2023), 1–24

    2023

  8. [16]

    Bowen Hao, Jing Zhang, Hongzhi Yin, Cuiping Li, and Hong Chen. 2021. Pre- training graph neural networks for cold-start users and items representation. In Proceedings of the 14th ACM International Conference on Web Search and Data Mining. 265–273

    2021

  9. [17]

    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

  10. [18]

    Po-Sen Huang, Xiaodong He, Jianfeng Gao, Li Deng, Alex Acero, and Larry Heck. 2013. Learning deep structured semantic models for web search using clickthrough data. In Proceedings of the 22nd ACM international conference on Information & Knowledge Management . 2333–2338

    2013

  11. [19]

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

    2018

  12. [20]

    James Kirkpatrick, Razvan Pascanu, Neil Rabinowitz, Joel Veness, Guillaume Desjardins, Andrei A Rusu, Kieran Milan, John Quan, Tiago Ramalho, Agnieszka Grabska-Barwinska, et al. 2017. Overcoming catastrophic forgetting in neural networks. Proceedings of the national academy of...

    2017

  13. [21]

    Brian Lester, Rami Al-Rfou, and Noah Constant. 2021. The power of scale for parameter-efficient prompt tuning. arXiv preprint arXiv:2104.08691 (2021)

    2021 arXiv

  14. [22]

    Lei Li, Yongfeng Zhang, and Li Chen. 2023. Personalized prompt learning for explainable recommendation. ACM Transactions on Information Systems 41, 4 (2023), 1–26

    2023

  15. [23]

    Xiang Lisa Li and Percy Liang. 2021. Prefix-tuning: Optimizing continuous prompts for generation. arXiv preprint arXiv:2101.00190 (2021)

    2021 arXiv

  16. [24]

    Taichi Liu, Chen Gao, Zhenyu Wang, Dong Li, Jianye Hao, Depeng Jin, and Yong Li. 2023. Uncertainty-aware Consistency Learning for Cold-Start Item Recommendation. In Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval ...

    2023

  17. [25]

    Xiao Liu, Kaixuan Ji, Yicheng Fu, Weng Tam, Zhengxiao Du, Zhilin Yang, and Jie Tang. 2022. P-tuning: Prompt tuning can be comparable to fine-tuning across scales and tasks. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Sh...

    2022

  18. [26]

    Yuanfu Lu, Yuan Fang, and Chuan Shi. 2020. Meta-learning on heterogeneous information networks for cold-start recommendation. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining . 1563–1573

    2020

  19. [27]

    Bonan Min, Hayley Ross, Elior Sulem, Amir Pouran Ben Veyseh, Thien Huu Nguyen, Oscar Sainz, Eneko Agirre, Ilana Heintz, and Dan Roth. 2023. Recent advances in natural language processing via large pre-trained language models: A survey. Comput. Surveys 56, 2 (2023), 1–40

    2023

  20. [28]

    Feiyang Pan, Shuokai Li, Xiang Ao, Pingzhong Tang, and Qing He. 2019. Warm up cold-start advertisements: Improving ctr predictions via learning to learn id embeddings. In Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retr...

    2019

  21. [29]

    Suvash Sedhain, Scott Sanner, Darius Braziunas, Lexing Xie, and Jordan Chris- tensen. 2014. Social collaborative filtering for cold-start recommendations. In Proceedings of the 8th ACM Conference on Recommender systems . 345–348

    2014

  22. [30]

    The TMall Team. [n. d.]. The Tmall Dataset. https://tianchi.aliyun.com/dataset/ 42

  23. [31]

    Aaron Van den Oord, Sander Dieleman, and Benjamin Schrauwen. 2013. Deep content-based music recommendation. Advances in neural information processing systems 26 (2013)

    2013

  24. [32]

    Laurens Van der Maaten and Geoffrey Hinton. 2008. Visualizing data using t-SNE. Journal of machine learning research 9, 11 (2008)

    2008

  25. [33]

    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

  26. [34]

    Maksims Volkovs, Guangwei Yu, and Tomi Poutanen. 2017. Dropoutnet: Ad- dressing cold start in recommender systems. Advances in neural information processing systems 30 (2017)

    2017

  27. [35]

    Ruoxi Wang, Bin Fu, Gang Fu, and Mingliang Wang. 2017. Deep & cross network for ad click predictions. In Proceedings of the ADKDD’17 . 1–7

    2017

  28. [36]

    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 ACM SIGKDD Conference on Knowledge Dis- covery and Data Mining . 1929–1937

    2022

  29. [37]

    Yuhao Wang, Xiangyu Zhao, Bo Chen, Qidong Liu, Huifeng Guo, Huanshuo Liu, Yichao Wang, Rui Zhang, and Ruiming Tang. 2023. PLATE: A Prompt- Enhanced Paradigm for Multi-Scenario Recommendations. In Proceedings of the 46th International ACM SIGIR Conference on Research and Develo...

    2023

  30. [38]

    Jian Wei, Jianhua He, Kai Chen, Yi Zhou, and Zuoyin Tang. 2017. Collaborative filtering and deep learning based recommendation system for cold start items. Expert Systems with Applications 69 (2017), 29–39

    2017

  31. [39]

    Xuansheng Wu, Huachi Zhou, Wenlin Yao, Xiao Huang, and Ninghao Liu. 2023. Towards Personalized Cold-Start Recommendation with Prompts. arXiv preprint arXiv:2306.17256 (2023)

    2023 arXiv

  32. [40]

    Yiqing Wu, Ruobing Xie, Yongchun Zhu, Fuzhen Zhuang, Xu Zhang, Leyu Lin, and Qing He. 2024. Personalized Prompt for Sequential Recommendation. IEEE Transactions on Knowledge and Data Engineering (2024)

    2024

  33. [41]

    Chaojun Xiao, Ruobing Xie, Yuan Yao, Zhiyuan Liu, Maosong Sun, Xu Zhang, and Leyu Lin. 2021. Uprec: User-aware pre-training for recommender systems. arXiv preprint arXiv:2102.10989 (2021)

    2021 arXiv

  34. [42]

    Lanling Xu, Junjie Zhang, Bingqian Li, Jinpeng Wang, Mingchen Cai, Wayne Xin Zhao, and Ji-Rong Wen. 2024. Prompting Large Language Models for Recom- mender Systems: A Comprehensive Framework and Empirical Analysis. arXiv preprint arXiv:2401.04997 (2024)

    2024 arXiv

  35. [43]

    Zixuan Yi, Iadh Ounis, and Craig Macdonald. 2023. Contrastive graph prompt- tuning for cross-domain recommendation. ACM Transactions on Information Systems 42, 2 (2023), 1–28

    2023

  36. [44]

    Jianyang Zhai, Xiawu Zheng, Chang-Dong Wang, Hui Li, and Yonghong Tian

  37. [45]

    Yin Zhang, Derek Zhiyuan Cheng, Tiansheng Yao, Xinyang Yi, Lichan Hong, and Ed H Chi. 2021. A model of two tales: Dual transfer learning framework for improved long-tail item recommendation. In Proceedings of the web conference

    2021

  38. [46]

    Yin Zhang, Ruoxi Wang, Derek Zhiyuan Cheng, Tiansheng Yao, Xinyang Yi, Lichan Hong, James Caverlee, and Ed H Chi. 2023. Empowering Long-tail Item Prompt Tuning for Item Cold-start Recommendation RecSys ’24, October 14–18, 2024, Bari, Italy Recommendation through Cross Decoupli...

    2023

  39. [47]

    Yongchun Zhu, Ruobing Xie, Fuzhen Zhuang, Kaikai Ge, Ying Sun, Xu Zhang, Leyu Lin, and Juan Cao. 2021. Learning to warm up cold item embeddings for cold- start recommendation with meta scaling and shifting networks. In Proceedings of the 44th International ACM SIGIR Conference...

    2021

  40. [2023]

    In Proceedings of the 31st ACM International Conference on Multimedia

    Knowledge Prompt-tuning for Sequential Recommendation. In Proceedings of the 31st ACM International Conference on Multimedia . 6451–6461

Pith tools

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