緩存允許將經常訪問的文件存儲在靠近用戶的邊緣設備上。文件的流行程度在緩存策略中起著至關重要的作用，協同過濾技術在塑造個性化推薦系統方面發揮了重要作用。在這項研究中，我們提出了一種將協同過濾與神經網絡結合的新方法，並以聯合多任務學習為支持。我們的模型預測用戶對項目的訪問概率，並基於文件的預期流行度、文件大小和其他系統因素做出緩存決策。用戶偏好和緩存管理的集成促使了一個動態且個性化的內容交付系統。通過使用聯邦學習，我們使每個用戶的偏好能夠影響緩存決策，從而提高了緩存內容的質量和相關性。使用實際數據集進行的廣泛實驗驗證了我們方法的有效性，顯示出系統效率的顯著提高、系統成本的優化以及整體用戶體驗的增強。該研究通過創新地將協同過濾、神經網絡和聯合多任務學習整合在一起，為內容協作緩存系統的發展做出了貢獻。

Caching allows the storage of frequently accessed files on edge devices near users. The popularity of files plays a crucial role in caching strategies, and collaborative filtering techniques have been instrumental in shaping personalized recommendation systems. In this study, we propose a novel approach that combines collaborative filtering with neural networks, supported by joint multi-task learning. Our model predicts user access probabilities for items and makes caching decisions based on the expected popularity of files, file size, and other system factors. The integration of user preferences and cache management facilitates a dynamic and personalized content delivery system. Through the use of federated learning, we enable each user's preferences to influence caching decisions, thereby enhancing the quality and relevance of cached content. Extensive experiments with real datasets validate the effectiveness of our approach, demonstrating significant improvements in system efficiency, optimized system cost, and overall enhanced user experience. This research contributes to the development of content collaboration caching systems through the innovative integration of collaborative filtering, neural networks, and joint multi-task learning.

Contents
Abstract-i
Content-ii
1 Introduction-1
2 Background and Related Works-4
2.1 Recommendation System-4
2.2 Distributed Learning-5
2.2.1 General Distributed Learning-5
2.2.2 Distributed Recommendation System-6
2.3 Caching Strategy-7
3 Multimedia Content Recommendation-9
3.1 Data Collection-9
3.2 Database Introduction-11
4 Caching by Multi-Task Federated Learning-12
4.1 System Model-12
4.2 Problem Formulation-14
4.2.1 Optimization-16
5 Learning Based Collaborative Filtering-19
5.1 Proposed Model-19
5.2 Federated Learning-20
5.2.1 Local Training-20
5.2.2 Aggregation-21
6 Experimental Results-22
6.1 Dataset Description-22
6.2 Baseline Methods-23
6.3 Parameter Settings-24
6.4 Dataset Configuration-24
6.5 Dataset Completion-25
6.6 Evaluation Protocols-25
6.7 Experiment Results-27
7 Conclusion-36
Bibliography-37

List of Figures
3.1 Scatter plot of user rating and AU of Happiness-11
4.1 System model-13
5.1 Proposed model-20
6.1 An example for cache entities positions-27
6.2 Mean square error for different models-28
6.3 Mean square error between true and predicted probability-29
6.4 System cost comparison for different models-32
6.7 System cost comparison for different caching methods-33
6.8 System cost comparison for different entities position-34
6.9 System cost comparison for different file sizes range-35

[1] K. Poularakis, G. Iosifidis, A. Argyriou, I. Koutsopoulos, and L. Tassiulas, “Caching and operator cooperation policies for layered video content delivery,” in IEEE INFOCOM 2016 - The 35th Annual IEEE International Conference on Computer Communications, (San Francisco, CA, USA), 2016.
[2] Y. Koren, R. Bell, and C. Volinsky, “Matrix factorization techniques for recommender systems,” Computer, vol. 42, no. 8, pp. 30–37, 2009.
[3] X. Wang, X. He, M. Wang, F. Feng, and T.-S. Chua, “Neural graph collaborative filtering,” in Proceedings of the 42nd international ACM SIGIR conference on Research and development in Information Retrieval, pp. 165–174, 2019.
[4] X. He, X. Du, X. Wang, F. Tian, J. Tang, and T.-S. Chua, “Outer product-based neural collaborative filtering,” arXiv preprint arXiv:1808.03912, 2018.
[5] H.-J. Xue, X. Dai, J. Zhang, S. Huang, and J. Chen, “Deep matrix factorization models for recommender systems.,” in IJCAI, vol. 17, pp. 3203–3209, Melbourne, Australia, 2017.
[6] X. He, L. Liao, H. Zhang, L. Nie, X. Hu, and T.-S. Chua, “Neural collaborative filtering,” in Proceedings of the 26th international conference on world wide web, pp. 173–182, 2017.
[7] H.-T. Cheng, L. Koc, J. Harmsen, T. Shaked, T. Chandra, H. Aradhye, G. Anderson, G. Corrado, W. Chai, M. Ispir, et al., “Wide & deep learning for recommender systems,” in Proceedings of the 1st workshop on deep learning for recommender systems, pp. 7–10, 2016.
[8] Z. Yu, J. Hu, G. Min, H. Lu, Z. Zhao, H. Wang, and N. Georgalas, “Federated learning based proactive content caching in edge computing,” in 2018 IEEE Global Communications Conference (GLOBECOM), pp. 1–6, IEEE, 2018.
[9] K. Qi and C. Yang, “Popularity prediction with federated learning for proactive caching at wireless edge,” in 2020 IEEE Wireless Communications and Networking Conference (WCNC), pp. 1–6, IEEE, 2020.
[10] B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. A. y Arcas, “Communicationefficient learning of deep networks from decentralized data,” in Artificial intelligence and statistics, pp. 1273–1282, PMLR, 2017.
[11] X. Ma, L. Zhao, G. Huang, Z. Wang, Z. Hu, X. Zhu, and K. Gai, “Entire space multitask model: An effective approach for estimating post-click conversion rate,” in The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval,
pp. 1137–1140, 2018.
[12] Z. Zhao, L. Hong, L. Wei, J. Chen, A. Nath, S. Andrews, A. Kumthekar, M. Sathiamoorthy, X. Yi, and E. Chi, “Recommending what video to watch next: a multitask ranking system,” in Proceedings of the 13th ACM Conference on Recommender Systems, pp. 43–51, 2019.
[13] H. Tang, J. Liu, M. Zhao, and X. Gong, “Progressive layered extraction (ple): A novel multitask learning (mtl) model for personalized recommendations,” in Fourteenth ACM Conference on Recommender Systems, pp. 269–278, 2020.
[14] J. Koneˇcn`y, H. B. McMahan, F. X. Yu, P. Richtárik, A. T. Suresh, and D. Bacon, “Federated learning: Strategies for improving communication efficiency,” arXiv preprint arXiv:1610.05492, 2016.
[15] V. Smith, C.-K. Chiang, M. Sanjabi, and A. S. Talwalkar, “Federated multi-task learning,” Advances in neural information processing systems, vol. 30, 2017.
[16] N. Bastianello, A. I. Rikos, and K. H. Johansson, “Online distributed learning with quantized finite-time coordination,” in 2023 62nd IEEE Conference on Decision and Control (CDC), IEEE, 2023.
[17] Z. Wang, X. Zhang, and Y. Chang, “Distributed sequential federated learning,” arXiv preprint arXiv:2302.00107, 2023.
[18] M. Ammad-Ud-Din, E. Ivannikova, S. A. Khan, W. Oyomno, Q. Fu, K. E. Tan, and A. Flanagan, “Federated collaborative filtering for privacy-preserving personalized recommendation system,” arXiv preprint arXiv:1901.09888, 2019.
[19] K. Muhammad, Q. Wang, D. O’Reilly-Morgan, E. Tragos, B. Smyth, N. Hurley, J. Geraci, and A. Lawlor, “Fedfast: Going beyond average for faster training of federated recommender systems,” in Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, pp. 1234–1242, 2020.
[20] H. Hazem, A. Awad, and A. Hassan, “A distributed real-time recommender system for big data streams,” arXiv preprint arXiv:2204.04633, 2022.
[21] T. Xiao, T. Cui, S. M. R. Islam, and Q. Chen, “Joint content placement and storage allocation based on federated learning in f-rans,” Sensors, vol. 21, 2021.
[22] Y. Wu, Y. Jiang, M. Bennis, F. Zheng, X. Gao, and X. You, “Content popularity prediction in fog radio access networks: A federated learning based approach,” in ICC 2020-2020 IEEE International Conference on Communications (ICC), pp. 1–6, IEEE, 2020.
[23] J. Ma, Z. Zhao, X. Yi, J. Chen, L. Hong, and E. H. Chi, “Modeling task relationships in multitask learning with multi-gate mixture-of-experts,” in Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, pp. 1930–1939, 2018.
[24] F. M. Harper and J. A. Konstan, “The movielens datasets: History and context,” Acm transactions on interactive intelligent systems (tiis), vol. 5, no. 4, pp. 1–19, 2015.
[25] B. M. Marlin, “Modeling user rating profiles for collaborative filtering,” Advances in neural information processing systems, vol. 16, 2003.