The Impact of Differential Privacy on Recommendation Accuracy and Popularity Bias | SpringerLink
Skip to main content
Springer Nature Link
Log in

The Impact of Differential Privacy on Recommendation Accuracy and Popularity Bias

  • Conference paper
  • First Online:
Advances in Information Retrieval (ECIR 2024)

Part of the book series: Lecture Notes in Computer Science ((LNCS,volume 14611))

Included in the following conference series:

Abstract

Collaborative filtering-based recommender systems leverage vast amounts of behavioral user data, which poses severe privacy risks. Thus, often random noise is added to the data to ensure Differential Privacy (DP). However, to date, it is not well understood in which ways this impacts personalized recommendations. In this work, we study how DP affects recommendation accuracy and popularity bias when applied to the training data of state-of-the-art recommendation models. Our findings are three-fold: First, we observe that nearly all users’ recommendations change when DP is applied. Second, recommendation accuracy drops substantially while recommended item popularity experiences a sharp increase, suggesting that popularity bias worsens. Finally, we find that DP exacerbates popularity bias more severely for users who prefer unpopular items than for users who prefer popular items.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Log in via an institution

Subscribe and save

Springer+ Basic
¥17,985 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Chapter
JPY 3498
Price includes VAT (Japan)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
JPY 13727
Price includes VAT (Japan)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
JPY 17159
Price includes VAT (Japan)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Similar content being viewed by others

Notes

  1. 1.

    The number of recommended relevant items is divided by the number of all relevant items (i.e., Recall), or by the length of the recommendation list (i.e., Precision). When DP is applied, \(\varDelta Recall\) and \(\varDelta Precision\) only depend on how the number of recommended relevant items changes and therefore, the relative change is the same.

  2. 2.

    https://github.com/pmuellner/ImpactOfDP.

  3. 3.

    No clear pattern across datasets can be observed [5] and thus, this behavior of MultVAE needs to be researched in the future.

References

  1. Abdollahpouri, H., et al.: Multistakeholder recommendation: survey and research directions. User Model. User-Adap. Inter. 30, 127–158 (2020)

    Article  Google Scholar 

  2. Abdollahpouri, H., Mansoury, M., Burke, R., Mobasher, B.: The unfairness of popularity bias in recommendation. In: Workshop on Recommendation in Multi-stakeholder Environments (RMSE), in Conjunction With the 13th ACM Conference on Recommender Systems (RecSys) (2019)

    Google Scholar 

  3. Abdollahpouri, H., Mansoury, M., Burke, R., Mobasher, B.: The connection between popularity bias, calibration, and fairness in recommendation. In: Proceedings of the 14th ACM Conference on Recommender Systems (RecSys), pp. 726–731 (2020)

    Google Scholar 

  4. Agarwal, S.: Trade-offs between fairness, interpretability, and privacy in machine learning. Master’s thesis, University of Waterloo (2020)

    Google Scholar 

  5. Anelli, V.W., Bellogín, A., Di Noia, T., Jannach, D., Pomo, C.: Top-n recommendation algorithms: a quest for the state-of-the-art. In: Proceedings of the 30th ACM Conference on User Modeling, Adaptation and Personalization (UMAP), pp. 121–131 (2022)

    Google Scholar 

  6. Bagdasaryan, E., Poursaeed, O., Shmatikov, V.: Differential privacy has disparate impact on model accuracy. In: Proceedings of the 33rd International Conference on Neural Information Processing Systems (NeurIPS), pp. 15479–15488 (2019)

    Google Scholar 

  7. Beigi, G., Liu, H.: A survey on privacy in social media: identification, mitigation, and applications. ACM Trans. Data Sci. (TDS) 1(1), 1–38 (2020)

    Article  Google Scholar 

  8. Berkovsky, S., Kuflik, T., Ricci, F.: The impact of data obfuscation on the accuracy of collaborative filtering. Expert Syst. Appl. 39(5), 5033–5042 (2012)

    Article  Google Scholar 

  9. Bishop, C.M.: Training with noise is equivalent to tikhonov regularization. Neural Comput. 7(1), 108–116 (1995)

    Article  Google Scholar 

  10. Calandrino, J.A., Kilzer, A., Narayanan, A., Felten, E.W., Shmatikov, V.: “you might also like:" privacy risks of collaborative filtering. In: 2011 IEEE Symposium on Security and Privacy (S &P), pp. 231–246 (2011)

    Google Scholar 

  11. Chen, C., Zhou, J., Wu, B., Fang, W., Wang, L., Qi, Y., Zheng, X.: Practical privacy preserving poi recommendation. ACM Trans. Intell. Syst. Technol. (TIST) 11(5), 1–20 (2020)

    Article  Google Scholar 

  12. Chen, C., Zhang, M., Zhang, Y., Liu, Y., Ma, S.: Efficient neural matrix factorization without sampling for recommendation. ACM Trans. Inf. Syst. (TOIS) 38(2), 1–28 (2020)

    Google Scholar 

  13. Ding, B., Kulkarni, J., Yekhanin, S.: Collecting telemetry data privately. In: Proceedings of the 31st International Conference on Neural Information Processing Systems (NeurIPS), pp. 3574–3583 (2017)

    Google Scholar 

  14. Dwork, C.: Differential privacy: a survey of results. In: International conference on Theory and Applications of Models of Computation (TAMC), pp. 1–19 (2008)

    Google Scholar 

  15. Dwork, C., Hardt, M., Pitassi, T., Reingold, O., Zemel, R.: Fairness through awareness. In: Proceedings of the 3rd Innovations in Theoretical Computer Science Conference (ITCS), pp. 214–226 (2012)

    Google Scholar 

  16. Dwork, C., Roth, A., et al.: The algorithmic foundations of differential privacy. Now Publishers, Inc. (2014)

    Google Scholar 

  17. Ekstrand, M.D., Joshaghani, R., Mehrpouyan, H.: Privacy for all: ensuring fair and equitable privacy protections. In: Proceedings of ACM Conference on Fairness, Accountability, and Transparency (FAccT), pp. 35–47 (2018)

    Google Scholar 

  18. Eskandanian, F., Sonboli, N., Mobasher, B.: Power of the few: analyzing the impact of influential users in collaborative recommender systems. In: Proceedings of the 27th ACM Conference on User Modeling, Adaptation and Personalization, pp. 225–233 (2019)

    Google Scholar 

  19. Friedman, A., Berkovsky, S., Kaafar, M.A.: A differential privacy framework for matrix factorization recommender systems. User Model. User-Adapt. Interact. (UMUAI) 26(5), 425–458 (2016)

    Article  Google Scholar 

  20. Friedman, A., Knijnenburg, B.P., Vanhecke, K., Martens, L., Berkovsky, S.: Privacy aspects of recommender systems. In: Ricci, F., Rokach, L., Shapira, B. (eds.) Recommender Systems Handbook, pp. 649–688. Springer, Boston, MA (2015). https://doi.org/10.1007/978-1-4899-7637-6_19

    Chapter  Google Scholar 

  21. Ganhör, C., Penz, D., Rekabsaz, N., Lesota, O., Schedl, M.: Unlearning protected user attributes in recommendations with adversarial training. In: Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR), pp. 2142–2147. Springer, Heidelberg (2022)

    Google Scholar 

  22. Gentry, C.: A fully homomorphic encryption scheme. Ph.D. thesis, Stanford university (2009)

    Google Scholar 

  23. Harper, F.M., Konstan, J.A.: The movielens datasets: history and context. ACM Trans. Interact. Intell. Syst. (TiiS) 5(4), 1–19 (2015)

    Google Scholar 

  24. Hashemi, H., et al.: Data leakage via access patterns of sparse features in deep learning-based recommendation systems. Workshop on Trustworthy and Socially Responsible Machine Learning (TSRML), in Conjunction with the 36th Conference on Neural Information Processing Systems (NeurIPS) (2022)

    Google Scholar 

  25. He, X., Deng, K., Wang, X., Li, Y., Zhang, Y., Wang, M.: Lightgcn: simplifying and powering graph convolution network for recommendation. In: Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR), pp. 639–648. Springer, Heidelberg (2020)

    Google Scholar 

  26. Kim, S., Kim, J., Koo, D., Kim, Y., Yoon, H., Shin, J.: Efficient privacy-preserving matrix factorization via fully homomorphic encryption. In: Proceedings of the 11th ACM on Asia Conference on Computer and Communications Security (ASIACCS), pp. 617–628 (2016)

    Google Scholar 

  27. Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. In: Proceedings of 3rd International Conference on Learning Representations (ICLR) (2015)

    Google Scholar 

  28. Klimashevskaia, A., Elahi, M., Jannach, D., Trattner, C., Skjærven, L.: Mitigating popularity bias in recommendation: potential and limits of calibration approaches. In: Advances in Information Retrieval: Workshop on Algorithmic Bias in Search and Recommendation (BIAS) in conjunction with the 42nd European Conference on IR Research (ECIR), pp. 82–90. Springer, Heidelberg (2022). https://doi.org/10.1007/978-3-031-09316-6_8

  29. Kowald, D., Schedl, M., Lex, E.: The unfairness of popularity bias in music recommendation: a reproducibility study. In: Jose, J.M., et al. (eds.) ECIR 2020. LNCS, vol. 12036, pp. 35–42. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-45442-5_5

    Chapter  Google Scholar 

  30. Lacic, E., Reiter-Haas, M., Kowald, D., Reddy Dareddy, M., Cho, J., Lex, E.: Using autoencoders for session-based job recommendations. User Model. User-Adap. Inter. 30, 617–658 (2020)

    Article  Google Scholar 

  31. Lesota, O., et al.: Analyzing item popularity bias of music recommender systems: are different genders equally affected? In: Proceedings of the 15th ACM Conference on Recommender Systems (RecSys), pp. 601–606 (2021)

    Google Scholar 

  32. Lex, E., Kowald, D., Schedl, M.: Modeling popularity and temporal drift of music genre preferences. Trans. Int. Soc. Music Inf. Retr. 3(1) (2020)

    Google Scholar 

  33. Liang, D., Krishnan, R.G., Hoffman, M.D., Jebara, T.: Variational autoencoders for collaborative filtering. In: Proceedings of the World Wide Web Conference (TheWebConf), pp. 689–698 (2018)

    Google Scholar 

  34. Lin, Y., et al.: Meta matrix factorization for federated rating predictions. In: Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR), pp. 981–990. Springer, Heidelberg (2020)

    Google Scholar 

  35. Long, J., Chen, T., Nguyen, Q.V.H., Yin, H.: Decentralized collaborative learning framework for next poi recommendation. ACM Trans. Inf. Syst. 41(3) (2023). https://doi.org/10.1145/3555374

  36. McMahan, B., Moore, E., Ramage, D., Hampson, S., Arcas, B.A.: Communication-efficient learning of deep networks from decentralized data. In: Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS), pp. 1273–1282 (2017)

    Google Scholar 

  37. Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., Galstyan, A.: A survey on bias and fairness in machine learning. ACM Comput. Surv. (CSUR) 54(6), 1–35 (2021)

    Article  Google Scholar 

  38. Melchiorre, A.B., Rekabsaz, N., Parada-Cabaleiro, E., Brandl, S., Lesota, O., Schedl, M.: Investigating gender fairness of recommendation algorithms in the music domain. Inf. Process. Manag. (IP &P) 58(5), 102666 (2021)

    Article  Google Scholar 

  39. Müllner, P., Lex, E., Schedl, M., Kowald, D.: Reuseknn: neighborhood reuse for differentially-private knn-based recommendations. ACM Trans. Intell. Syst. Technol. (2023). https://doi.org/10.1145/3608481

    Article  Google Scholar 

  40. Muellner, P., Kowald, D., Lex, E.: Robustness of meta matrix factorization against strict privacy constraints. In: Hiemstra, D., Moens, M.-F., Mothe, J., Perego, R., Potthast, M., Sebastiani, F. (eds.) ECIR 2021. LNCS, vol. 12657, pp. 107–119. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-72240-1_8

    Chapter  Google Scholar 

  41. Müllner, P., Lex, E., Schedl, M., Kowald, D.: Differential privacy in collaborative filtering recommender systems: a review. Front. Big Data 6 (2023). https://doi.org/10.3389/fdata.2023.1249997

  42. Nasr, M., Shokri, R., Houmansadr, A.: Comprehensive privacy analysis of deep learning: Passive and active white-box inference attacks against centralized and federated learning. In: Proceedings of the IEEE Symposium on Security and Privacy (S &P), pp. 739–753 (2019)

    Google Scholar 

  43. Ni, J., Li, J., McAuley, J.: Justifying recommendations using distantly-labeled reviews and fine-grained aspects. In: Proceedings of the Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pp. 188–197 (2019)

    Google Scholar 

  44. Parra, D., Sahebi, S.: Recommender systems: sources of knowledge and evaluation metrics. In: Advanced Techniques in Web Intelligence-2: Web User Browsing Behaviour and Preference Analysis, pp. 149–175. Springer, Heidelberg (2013). https://doi.org/10.1007/978-3-642-33326-2_7

  45. Ren, H., Deng, J., Xie, X.: GRNN: generative regression neural network-a data leakage attack for federated learning. ACM Trans. Intell. Syst. Technol. (TIST) 13(4), 1–24 (2022)

    Google Scholar 

  46. Saveski, M., Mantrach, A.: Item cold-start recommendations: learning local collective embeddings. In: Proceedings of the 8th ACM Conference on Recommender systems (RecSys), pp. 89–96 (2014)

    Google Scholar 

  47. Schedl, M., Bauer, C.: Distance-and rank-based music mainstreaminess measurement. In: Adjunct Publication of the 25th Conference on User Modeling, Adaptation and Personalization (UMAP): Workshop on Surprise, Opposition, and Obstruction in Adaptive and Personalized Systems (SOAP), pp. 364–367 (2017)

    Google Scholar 

  48. Schedl, M., Bauer, C., Reisinger, W., Kowald, D., Lex, E.: Listener modeling and context-aware music recommendation based on country archetypes. Front. Artif. Intell. 3, 508725 (2021)

    Article  Google Scholar 

  49. Lam, S.K., Frankowski, D., Riedl, J.: Do you trust your recommendations? an exploration of security and privacy issues in recommender systems. In: Müller, G. (ed.) ETRICS 2006. LNCS, vol. 3995, pp. 14–29. Springer, Heidelberg (2006). https://doi.org/10.1007/11766155_2

    Chapter  Google Scholar 

  50. Sun, J.A., Pentyala, S., Cock, M.D., Farnadi, G.: Privacy-preserving fair item ranking. In: Kamps, J., et al. (eds.) ECIR 2023, vol. 13981, pp. 188–203. Springer, Heidelberg (2023). https://doi.org/10.1007/978-3-031-28238-6_13

    Chapter  Google Scholar 

  51. Sun, Z., et al.: Are we evaluating rigorously? benchmarking recommendation for reproducible evaluation and fair comparison. In: Proceedings of the 14th ACM Conference on Recommender Systems (RecSys), pp. 23–32 (2020)

    Google Scholar 

  52. Weinsberg, U., Bhagat, S., Ioannidis, S., Taft, N.: Blurme: inferring and obfuscating user gender based on ratings. In: Proceedings of the 6th ACM Conference on Recommender Systems (RecSys), pp. 195–202 (2012)

    Google Scholar 

  53. Xin, X., et al.: On the user behavior leakage from recommender system exposure. ACM Trans. Inf. Syst. (TOIS) 41(3), 1–25 (2023)

    Article  Google Scholar 

  54. Xin, Y., Jaakkola, T.: Controlling privacy in recommender systems. In: Proceedings of the 27th International Conference on Neural Information Processing Systems (NeurIPS), pp. 2618–2626. MIT Press, Cambridge (2014)

    Google Scholar 

  55. Yang, Z., Ge, Y., Su, C., Wang, D., Zhao, X., Ying, Y.: Fairness-aware differentially private collaborative filtering. In: Companion Proceedings of the ACM Web Conference (TheWebConf), pp. 927–931 (2023)

    Google Scholar 

  56. Zemel, R., Wu, Y., Swersky, K., Pitassi, T., Dwork, C.: Learning fair representations. In: International conference on machine learning (ICML), pp. 325–333 (2013)

    Google Scholar 

  57. Zhang, M., et al.: Membership inference attacks against recommender systems. In: Proceedings of the ACM SIGSAC Conference on Computer and Communications Security (CCS), pp. 864–879 (2021)

    Google Scholar 

  58. Zhang, S., Yin, H.: Comprehensive privacy analysis on federated recommender system against attribute inference attacks. IEEE Trans. Knowl. Data Eng. (TKDE) (2023)

    Google Scholar 

  59. Zhu, T., Li, G., Ren, Y., Zhou, W., Xiong, P.: Differential privacy for neighborhood-based collaborative filtering. In: Proceedings of the IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), pp. 752–759 (2013)

    Google Scholar 

Download references

Acknowledgments

This research is funded by the “DDAI” COMET Module within the COMET - Competence Centers for Excellent Technologies Programme, funded by the Austrian Federal Ministry for Transport, Innovation and Technology (bmvit), the Austrian Federal Ministry for Digital and Economic Affairs (bmdw), the Austrian Research Promotion Agency (FFG), the province of Styria (SFG) and partners from industry and academia. The COMET Programme is managed by FFG. Moreover, this research received support by the Austrian Science Fund (FWF): DFH-23 and P36413; and by the State of Upper Austria and the Federal Ministry of Education, Science, and Research, through grants LIT-2020-9-SEE-113 and LIT-2021-10-YOU-215. For open access purposes, the author has applied a CC BY public copyright license to any author accepted manuscript version arising from this submission.

Author information

Authors and Affiliations

  1. Know-Center GmbH, Graz, Austria

    Peter Müllner & Dominik Kowald

  2. Graz University of Technology, Graz, Austria

    Elisabeth Lex & Dominik Kowald

  3. Johannes Kepler University Linz, Linz, Austria

    Markus Schedl

  4. Linz Institute of Technology, Linz, Austria

    Markus Schedl

Authors
  1. Peter Müllner
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Elisabeth Lex
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Markus Schedl
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Dominik Kowald
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Peter Müllner .

Editor information

Editors and Affiliations

  1. Georgetown University, Washington, WA, USA

    Nazli Goharian

  2. University of Pisa, PISA, Pisa, Italy

    Nicola Tonellotto

  3. King's College London, London, UK

    Yulan He

  4. University College London, London, UK

    Aldo Lipani

  5. University of Glasgow, Glasgow, UK

    Graham McDonald

  6. University of Glasgow, Glasgow, UK

    Craig Macdonald

  7. University of Glasgow, Glasgow, UK

    Iadh Ounis

Rights and permissions

Reprints and permissions

Copyright information

© 2024 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Müllner, P., Lex, E., Schedl, M., Kowald, D. (2024). The Impact of Differential Privacy on Recommendation Accuracy and Popularity Bias. In: Goharian, N., et al. Advances in Information Retrieval. ECIR 2024. Lecture Notes in Computer Science, vol 14611. Springer, Cham. https://doi.org/10.1007/978-3-031-56066-8_33

Download citation

  • DOI: https://doi.org/10.1007/978-3-031-56066-8_33

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-031-56065-1

  • Online ISBN: 978-3-031-56066-8

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics

Search

Navigation