Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
102 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

On-Device Recommender Systems: A Comprehensive Survey (2401.11441v2)

Published 21 Jan 2024 in cs.IR

Abstract: Recommender systems have been widely deployed in various real-world applications to help users identify content of interest from massive amounts of information. Traditional recommender systems work by collecting user-item interaction data in a cloud-based data center and training a centralized model to perform the recommendation service. However, such cloud-based recommender systems (CloudRSs) inevitably suffer from excessive resource consumption, response latency, as well as privacy and security risks concerning both data and models. Recently, driven by the advances in storage, communication, and computation capabilities of edge devices, there has been a shift of focus from CloudRSs to on-device recommender systems (DeviceRSs), which leverage the capabilities of edge devices to minimize centralized data storage requirements, reduce the response latency caused by communication overheads, and enhance user privacy and security by localizing data processing and model training. Despite the rapid rise of DeviceRSs, there is a clear absence of timely literature reviews that systematically introduce, categorize and contrast these methods. To bridge this gap, we aim to provide a comprehensive survey of DeviceRSs, covering three main aspects: (1) the deployment and inference of DeviceRSs (2) the training and update of DeviceRSs (3) the security and privacy of DeviceRSs. Furthermore, we provide a fine-grained and systematic taxonomy of the methods involved in each aspect, followed by a discussion regarding challenges and future research directions. This is the first comprehensive survey on DeviceRSs that covers a spectrum of tasks to fit various needs. We believe this survey will help readers effectively grasp the current research status in this field, equip them with relevant technical foundations, and stimulate new research ideas for developing DeviceRSs.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (179)
  1. Improved Algorithms for Linear Stochastic Bandits. In NeurIPS. 2312–2320.
  2. Shipra Agrawal and Navin Goyal. 2013. Thompson sampling for contextual bandits with linear payoffs. In ICML. 127–135.
  3. Federated recommenders: methods, challenges and future. Cluster Computing 25, 6 (2022), 4075–4096.
  4. Federated collaborative filtering for privacy-preserving personalized recommendation system. arXiv preprint arXiv:1901.09888 (2019).
  5. NRDL: Decentralized user preference learning for privacy-preserving next POI recommendation. Expert Systems with Applications 239 (2024), 122421.
  6. Federank: User controlled feedback with federated recommender systems. In ECIR. 32–47.
  7. PIR with compressed queries and amortized query processing. In SP. 962–979.
  8. Felix Beierle and Tobias Eichinger. 2019. Collaborating with users in proximity for decentralized mobile recommender systems. In SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI. 1192–1197.
  9. Privacy-aware recommendation with private-attribute protection using adversarial learning. In WSDM. 34–42.
  10. Machine learning with adversaries: Byzantine tolerant gradient descent. In NeurIPS. 119–129.
  11. Practical Secure Aggregation for Privacy-Preserving Machine Learning. In SIGSAC. 1175–1191.
  12. Privacy Matters: Vertical Federated Linear Contextual Bandits for Privacy Protected Recommendation. In KDD. 154–166.
  13. Secure federated matrix factorization. IEEE Intelligent Systems 36, 5 (2020), 11–20.
  14. Efficient Federated Matrix Factorization Against Inference Attacks. ACM Trans. Intell. Syst. Technol. 13, 4 (2022), 1–20.
  15. Secure Social Recommendation Based on Secret Sharing. In ECAI. 506–512.
  16. Privacy preserving point-of-interest recommendation using decentralized matrix factorization. In AAAI. 257–264.
  17. Differential private knowledge transfer for privacy-preserving cross-domain recommendation. In WWW. 1455–1465.
  18. Robust federated recommendation system. arXiv preprint arXiv:2006.08259 (2020).
  19. Learning elastic embeddings for customizing on-device recommenders. In SIGKDD. 138–147.
  20. Differentiable neural input search for recommender systems. arXiv preprint arXiv:2006.04466 (2020).
  21. Konstantina Christakopoulou and Arindam Banerjee. 2019. Adversarial attacks on an oblivious recommender. In RecSys. 322–330.
  22. Binaryconnect: Training deep neural networks with binary weights during propagations. In NeurIPS. 3123–3131.
  23. Deep neural networks for youtube recommendations. In RecSys. 191–198.
  24. Exploiting data sparsity in secure cross-platform social recommendation. In NeurIPS. 10524–10534.
  25. M6-rec: Generative pretrained language models are open-ended recommender systems. arXiv preprint arXiv:2205.08084 (2022).
  26. A decentralized recommendation engine in the social internet of things. In UMAP. 77–82.
  27. A human-centered decentralized architecture and recommendation engine in SIoT. User Modeling and User-Adapted Interaction 32, 3 (2022), 297–353.
  28. DeepLight: Deep Lightweight Feature Interactions for Accelerating CTR Predictions in Ad Serving. In WSDM. 922–930.
  29. Heterogeneous federated collaborative filtering using FAIR: Federated Averaging in Random Subspaces. arXiv preprint arXiv:2311.01722 (2023).
  30. A Survey of On-Device Machine Learning: An Algorithms and Learning Theory Perspective. ACM Trans. Internet Things 2, 3 (2021), 49 pages.
  31. Taamr: Targeted adversarial attack against multimedia recommender systems. In DSN-W. 1–8.
  32. A survey of vision-language pre-trained models. arXiv preprint arXiv:2202.10936 (2022).
  33. Cynthia Dwork. 2008. Differential privacy: A survey of results. In TAMC. 1–19.
  34. Calibrating noise to sensitivity in private data analysis. In TCC. 265–284.
  35. Personalized federated learning with theoretical guarantees: A model-agnostic meta-learning approach. In NeurIPS. 3557–3568.
  36. Recommender systems in the era of large language models (llms). arXiv preprint arXiv:2307.02046 (2023).
  37. Poisoning attacks to graph-based recommender systems. In ACSAC. 381–392.
  38. Graph Neural Networks for Recommender System. In WSDM. 1623–1625.
  39. Craig Gentry. 2009. Fully homomorphic encryption using ideal lattices. In STOC. 169–178.
  40. Mixed dimension embeddings with application to memory-efficient recommendation systems. In ISIT. 2786–2791.
  41. Real-time Short Video Recommendation on Mobile Devices. In CIKM. 3103–3112.
  42. EdgeRec: recommender system on edge in Mobile Taobao. In CIKM. 2477–2484.
  43. Post-training 4-bit quantization on embedding tables. arXiv preprint arXiv:1911.02079 (2019).
  44. The hidden vulnerability of distributed learning in byzantium. In ICML. 3521–3530.
  45. Shilling attacks against recommender systems: a comprehensive survey. Artificial Intelligence Review 42 (2014), 767–799.
  46. DeepFM: a factorization-machine based neural network for CTR prediction. arXiv preprint arXiv:1703.04247 (2017).
  47. Understanding the scope and impact of the california consumer privacy act of 2018. Journal of Data Protection & Privacy 2, 3 (2019), 234–253.
  48. Xiangnan He and Tat-Seng Chua. 2017. Neural factorization machines for sparse predictive analytics. In SIGIR. 355–364.
  49. Neural collaborative filtering. In WWW. 173–182.
  50. Large language models as zero-shot conversational recommenders. In CIKM. 720–730.
  51. Decentralized recommendation based on matrix factorization: A comparison of gossip and federated learning. In ECML PKDD. 317–332.
  52. Balázs Hidasi and Alexandros Karatzoglou. 2018. Recurrent neural networks with top-k gains for session-based recommendations. In CIKM. 843–852.
  53. FedDSR: Daily Schedule Recommendation in a Federated Deep Reinforcement Learning Framework. IEEE Transactions on Knowledge and Data Engineering 35, 4 (2023), 3912–3924.
  54. ReFRS: Resource-Efficient Federated Recommender System for Dynamic and Diversified User Preferences. ACM Trans. Inf. Syst. 41, 3 (2023), 30 pages.
  55. Learning private neural language modeling with attentive aggregation. In IJCNN. 1–8.
  56. On the detection of shilling attacks in federated collaborative filtering. In SRDS. 185–194.
  57. Neural input search for large scale recommendation models. In SIGKDD. 2387–2397.
  58. Learning to embed categorical features without embedding tables for recommendation. In SIGKDD. 840–850.
  59. Wang-Cheng Kang and Julian McAuley. 2019. Candidate generation with binary codes for large-scale top-n recommendation. In CIKM. 1523–1532.
  60. A review of attacks and its detection attributes on collaborative recommender systems. International Journal of Advanced Research in Computer Science 8, 7 (2017).
  61. Application of random walks to decentralized recommender systems. In OPODIS. 48–63.
  62. A Payload Optimization Method for Federated Recommender Systems. In RecSys. 432–442.
  63. Matrix factorization techniques for recommender systems. Computer 42, 8 (2009), 30–37.
  64. Rating prediction based on social sentiment from textual reviews. IEEE transactions on multimedia 18, 9 (2016), 1910–1921.
  65. An algorithm for efficient privacy-preserving item-based collaborative filtering. Future Generation Computer Systems 55 (2016), 311–320.
  66. Adaptive low-precision training for embeddings in click-through rate prediction. In AAAI. 4435–4443.
  67. Embedding Compression in Recommender Systems: A Survey. Comput. Surveys 56, 5 (2023), 21 pages.
  68. Discrete content-aware matrix factorization. In SIGKDD. 325–334.
  69. LightRec: A Memory and Search-Efficient Recommender System. In www. 695–705.
  70. Fedrec++: Lossless federated recommendation with explicit feedback. In AAAI. 4224–4231.
  71. Attacking recommender systems with augmented user profiles. In CIKM. 855–864.
  72. Fedrec: Federated recommendation with explicit feedback. IEEE Intelligent Systems 36, 5 (2020), 21–30.
  73. Meta Matrix Factorization for Federated Rating Predictions. In SIGIR. 981–990.
  74. FR-FMSS: Federated Recommendation via Fake Marks and Secret Sharing. In RecSys. 668–673.
  75. A Generic Federated Recommendation Framework via Fake Marks and Secret Sharing. ACM Trans. Inf. Syst. 41, 2 (2022), 37 pages.
  76. Decentralized low-rank matrix completion. In ICASSP. 2925–2928.
  77. Automated Embedding Size Search in Deep Recommender Systems. In SIGIR. 2307–2316.
  78. Is chatgpt a good recommender? a preliminary study. arXiv preprint arXiv:2304.10149 (2023).
  79. PrivateRec: Differentially Private Model Training and Online Serving for Federated News Recommendation. In SIGKDD. 4539–4548.
  80. Learnable embedding sizes for recommender systems. arXiv preprint arXiv:2101.07577.
  81. Fairness-aware federated matrix factorization. In RecSys. 168–178.
  82. FedCT: Federated Collaborative Transfer for Recommendation. In SIGIR. 716–725.
  83. Federated Probabilistic Preference Distribution Modelling with Compactness Co-Clustering for Privacy-Preserving Multi-Domain Recommendation. In IJCAI. 2206–2214.
  84. Collaborative hashing. In CVPR. 2139–2146.
  85. Federated Social Recommendation with Graph Neural Network. ACM Trans. Intell. Syst. Technol. 13, 4 (2022), 24 pages.
  86. Model-Agnostic Decentralized Collaborative Learning for On-Device POI Recommendation. In SIGIR. 423–432.
  87. Decentralized collaborative learning framework for next POI recommendation. ACM Transactions on Information Systems 41, 3 (2023), 1–25.
  88. Personalized Federated Recommendation via Joint Representation Learning, User Clustering, and Model Adaptation. In CIKM. 4289–4293.
  89. PerFedRec++: Enhancing Personalized Federated Recommendation with Self-Supervised Pre-Training. arXiv preprint arXiv:2305.06622 (2023).
  90. OptEmbed: Learning Optimal Embedding Table for Click-through Rate Prediction. In CIKM. 1399–1409.
  91. Recommender systems with social regularization. In WSDM. 287–296.
  92. Towards fair federated recommendation learning: Characterizing the inter-dependence of system and data heterogeneity. In RecSys. 156–167.
  93. Improving activity data collection with on-device personalization using fine-tuning. In UbiComp/ISWC. 255–260.
  94. Communication-Efficient Learning of Deep Networks from Decentralized Data. In AISTATS. 1273–1282.
  95. FedCDR: Federated Cross-Domain Recommendation for Privacy-Preserving Rating Prediction. In CIKM. 2179–2188.
  96. Exploiting Unintended Feature Leakage in Collaborative Learning. In SP. 691–706.
  97. Recent advances in natural language processing via large pre-trained language models: A survey. Comput. Surveys 56, 2 (2023), 1–40.
  98. Stronger privacy for federated collaborative filtering with implicit feedback. In RecSys. 342–350.
  99. FedFast: Going Beyond Average for Faster Training of Federated Recommender Systems. In SIGKDD. 1234–1242.
  100. On first-order meta-learning algorithms. arXiv preprint arXiv:1803.02999 (2018).
  101. Learning compressed embeddings for on-device inference. Proceedings of Machine Learning and Systems 4 (2022), 382–397.
  102. Yoon-Joo Park and Alexander Tuzhilin. 2008. The long tail of recommender systems and how to leverage it. In RecSys. 11–18.
  103. FedPOIRec: Privacy-preserving federated poi recommendation with social influence. Information Sciences 623 (2023), 767–790.
  104. Vasileios Perifanis and Pavlos S. Efraimidis. 2022. Federated Neural Collaborative Filtering. Know.-Based Syst. 242, C (2022), 16 pages.
  105. Privacy-preserving news recommendation model learning. arXiv preprint arXiv:2003.09592 (2020).
  106. Semi-Decentralized Federated Ego Graph Learning for Recommendation. In WWW. 339–348.
  107. Single-shot embedding dimension search in recommender system. In SIGIR. 513–522.
  108. Continuous Input Embedding Size Search For Recommender Systems. In SIGIR. 708–717.
  109. Adaptive federated optimization. arXiv preprint arXiv:2003.00295 (2020).
  110. Poisoning deep learning based recommender model in federated learning scenarios. arXiv preprint arXiv:2204.13594 (2022).
  111. FedRecAttack: model poisoning attack to federated recommendation. In ICDE. 2643–2655.
  112. Adi Shamir. 1979. How to share a secret. Commun. ACM 22, 11 (1979), 612–613.
  113. Compositional embeddings using complementary partitions for memory-efficient recommendation systems. In SIGKDD. 165–175.
  114. A Survey on Federated Recommendation Systems. arXiv preprint arXiv:2301.00767 (2022).
  115. Personalized federated learning with moreau envelopes. In NeurIPS. 21394–21405.
  116. Learning to hash with graph neural networks for recommender systems. In WWW. 1988–1998.
  117. AutoML for Deep Recommender Systems: Fundamentals and Advances. In WSDM. 1264–1267.
  118. Graph attention networks. arXiv preprint arXiv:1710.10903 (2017).
  119. FedRec+: Enhancing Privacy and Addressing Heterogeneity in Federated Recommendation Systems. arXiv preprint arXiv:2310.20193 (2023).
  120. Next Point-of-Interest Recommendation on Resource-Constrained Mobile Devices. In WWW. 906–916.
  121. Fast-Adapting and Privacy-Preserving Federated Recommender System. The VLDB Journal 31, 5 (2021), 877–896.
  122. Shiqiang Wang and Mingyue Ji. 2022. A unified analysis of federated learning with arbitrary client participation. In NeurIPS. 19124–19137.
  123. Decentralized recommender systems. arXiv preprint arXiv:1503.01647 (2015).
  124. Llmrec: Large language models with graph augmentation for recommendation. arXiv preprint arXiv:2311.00423 (2023).
  125. Triple adversarial learning for influence based poisoning attack in recommender systems. In SIGKDD. 1830–1840.
  126. Fedgnn: Federated graph neural network for privacy-preserving recommendation. arXiv preprint arXiv:2102.04925 (2021).
  127. FedCTR: Federated Native Ad CTR Prediction with Cross-Platform User Behavior Data. ACM Trans. Intell. Syst. Technol. 13, 4 (2022), 19 pages.
  128. A federated graph neural network framework for privacy-preserving personalization. Nature Communications 13, 1 (2022), 3091.
  129. FedAttack: Effective and covert poisoning attack on federated recommendation via hard sampling. In SIGKDD. 4164–4172.
  130. Hierarchical Personalized Federated Learning for User Modeling. In WWW. 957–968.
  131. A survey on accuracy-oriented neural recommendation: From collaborative filtering to information-rich recommendation. IEEE Transactions on Knowledge and Data Engineering 35, 5 (2022), 4425–4445.
  132. Graph Neural Networks in Recommender Systems: A Survey. ACM Comput. Surv. 55, 5 (2022), 37 pages.
  133. HySAD: A semi-supervised hybrid shilling attack detector for trustworthy product recommendation. In SIGKDD. 985–993.
  134. Poisoning attacks against knowledge graph-based recommendation systems using deep reinforcement learning. Neural Computing and Applications (2022), 1–19.
  135. On-Device Next-Item Recommendation with Self-Supervised Knowledge Distillation. In SIGIR. 546–555.
  136. Efficient on-device session-based recommendation. ACM Transactions on Information Systems 41, 4 (2023), 1–24.
  137. Towards Communication-Efficient Model Updating for On-Device Session-Based Recommendation. In CIKM. 2795–2804.
  138. Agile and accurate CTR prediction model training for massive-scale online advertising systems. In SIGMOD. 2404–2409.
  139. Hard negative examples are hard, but useful. In ECCV. 126–142.
  140. Binary Code Based Hash Embedding for Web-Scale Applications. In CIKM. 3563–3567.
  141. Learning Effective and Efficient Embedding via an Adaptively-Masked Twins-based Layer. arXiv preprint arXiv:2108.11513 (2021).
  142. On-Device Learning for Model Personalization with Large-Scale Cloud-Coordinated Domain Adaption. In SIGKDD. 2180–2190.
  143. FCMF: Federated collective matrix factorization for heterogeneous collaborative filtering. Knowledge-Based Systems 220 (2021), 106946.
  144. Federated recommendation systems. Federated Learning: Privacy and Incentive (2020), 225–239.
  145. DPMF: Decentralized Probabilistic Matrix Factorization for Privacy-Preserving Recommendation. Applied Sciences 12, 21 (2022), 11118.
  146. Device-cloud collaborative learning for recommendation. In SIGKDD. 3865–3874.
  147. Tt-rec: Tensor train compression for deep learning recommendation models. Proceedings of Machine Learning and Systems 3 (2021), 448–462.
  148. Byzantine-robust distributed learning: Towards optimal statistical rates. In ICML. 5650–5659.
  149. Joint Modeling of Users’ Interests and Mobility Patterns for Point-of-Interest Recommendation. In MM. 819–822.
  150. Joint Modeling of User Check-in Behaviors for Real-time Point-of-Interest Recommendation. ACM Trans. Inf. Syst. 35, 2 (2016), 44 pages.
  151. See through gradients: Image batch recovery via gradinversion. In CVPR. 16337–16346.
  152. LCARS: a location-content-aware recommender system. In SIGKDD. 221–229.
  153. Spatial-Aware Hierarchical Collaborative Deep Learning for POI Recommendation. IEEE Transactions on Knowledge and Data Engineering 29, 11 (2017), 2537–2551.
  154. Graph convolutional neural networks for web-scale recommender systems. In SIGKDD. 974–983.
  155. Senci Ying. 2020. Shared MF: A privacy-preserving recommendation system. arXiv preprint arXiv:2008.07759 (2020).
  156. Self-supervised learning for recommender systems: A survey. arXiv preprint arXiv:2203.15876 (2022).
  157. Untargeted attack against federated recommendation systems via poisonous item embeddings and the defense. In AAAI. 4854–4863.
  158. Manipulating Federated Recommender Systems: Poisoning with Synthetic Users and Its Countermeasures. SIGIR (2023), 1690–1699.
  159. HeteFedRec: Federated Recommender Systems with Model Heterogeneity. arXiv preprint arXiv:2307.12810 (2023).
  160. Interaction-level Membership Inference Attack Against Federated Recommender Systems. In WWW. 1053–1062.
  161. Hide Your Model: A Parameter Transmission-free Federated Recommender System. arXiv preprint arXiv:2311.14968 (2023).
  162. Federated unlearning for on-device recommendation. In WSDM. 393–401.
  163. Manipulating Visually-aware Federated Recommender Systems and Its Countermeasures. ACM Transactions on Information Systems 42, 3 (2023), 1–26.
  164. Dual Personalization on Federated Recommendation. arXiv preprint arXiv:2301.08143 (2023).
  165. LightFR: Lightweight federated recommendation with privacy-preserving matrix factorization. ACM Transactions on Information Systems 41, 4 (2023), 1–28.
  166. Discrete collaborative filtering. In SIGIR. 325–334.
  167. Graph embedding for recommendation against attribute inference attacks. In WWW. 3002–3014.
  168. Pipattack: Poisoning federated recommender systems for manipulating item promotion. In WSDM. 1415–1423.
  169. Comprehensive privacy analysis on federated recommender system against attribute inference attacks. IEEE Transactions on Knowledge and Data Engineering (2023), 1–13.
  170. Discrete personalized ranking for fast collaborative filtering from implicit feedback. In AAAI, Vol. 31.
  171. Preference preserving hashing for efficient recommendation. In SIGIR. 183–192.
  172. AutoDim: Field-aware Embedding Dimension Searchin Recommender Systems. In WWW. 3015–3022.
  173. AutoEmb: Automated Embedding Dimensionality Search in Streaming Recommendations. arXiv preprint arXiv:2002.11252 (2020).
  174. AutoML for Deep Recommender Systems: A Survey. ACM Transactions on Information Systems 41, 4 (2023), 38 pages.
  175. Decentralized graph neural network for privacy-preserving recommendation. In CIKM. 3494–3504.
  176. Deep Interest Network for Click-Through Rate Prediction. In SIGKDD. 1059–1068.
  177. On-Device Learning Systems for Edge Intelligence: A Software and Hardware Synergy Perspective. IEEE Internet of Things Journal 8, 15 (2021), 11916–11934.
  178. Deep leakage from gradients. In NeurIPS. 14774–14784.
  179. Barret Zoph and Quoc V Le. 2016. Neural architecture search with reinforcement learning. arXiv preprint arXiv:1611.01578 (2016).
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (9)
  1. Hongzhi Yin (210 papers)
  2. Liang Qu (22 papers)
  3. Tong Chen (200 papers)
  4. Wei Yuan (110 papers)
  5. Ruiqi Zheng (11 papers)
  6. Jing Long (13 papers)
  7. Xin Xia (171 papers)
  8. Yuhui Shi (44 papers)
  9. Chengqi Zhang (74 papers)
Citations (24)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now