Using negative user actions to improve the quality of recommender systems

Cover Page

Cite item

Full Text

Open Access Open Access
Restricted Access Access granted
Restricted Access Subscription or Fee Access

Abstract

Recommendation systems are finding increasingly wide application, encompassing a variety of domains and diverse data types. However, in scenarios with a limited number of items, traditional approaches often prove to be insufficiently effective. In such cases, methods based on boosting algorithms offer a more efficient solution. This paper proposes a way to improve recommendation quality within this approach by incorporating users’ negative interactions with items. Integrating these data enables more accurate modeling of both preferences and avoidances. The presented method enhances recommendation personalization even under conditions of high interdependence and limited item availability.

Full Text

Restricted Access

About the authors

M. А. Zharova

Moscow Institute of Physics and Technology (MIPT); Federal Research Center “Computer Science and Control”, RAS

Author for correspondence.
Email: zharova.ma@phystech.edu
Russian Federation, Dolgoprudny; Moscow

V. I. Tsurkov

Federal Research Center “Computer Science and Control”, RAS

Email: v.tsurkov@frccsc.ru
Russian Federation, Moscow

References

  1. Cano E., Morisio M. Hybrid Recommender Systems: A Systematic Literature Review // Intelligent Data Analysis. 2017. V. 21. P. 1487–1524.
  2. Zharova M., Tsurkov V. Neural Network Approaches for Recommender Systems // J. Computer and Systems Sciences International. 2024. V. 62. P. 1048–1062.
  3. Zharova M., Tsurkov V. Boosting Based Recommender System // J. Computer and Systems Sciences International. 2024. V. 63. P. 922–940.
  4. Xinran H., Junfeng P., Ou J. Practical Lessons from Predicting Clicks on Ads at Facebook // Proc. 8th Intern. Workshop on Data Mining for Online Advertising. N.Y., USA, 2014. P 1–9.
  5. Paul C., Jay A., Emre S. Deep Neural Networks for YouTube Recommendations // Proc. 10th ACM Conf. on Recommender Systems. Boston, USA, 2016. P 191–198.
  6. Clark J. Target Variable Engineering // arXiv:2310.09440, 2023.
  7. Tschalzev A., Marton S. A Data-Centric Perspective on Evaluating Machine Learning Models for Tabular Data // arXiv:2407.02112, 2024.
  8. Ke1 G., Meng Q., Finley T. LightGBM: A Highly Efficient Gradient Boosting Decision Tree // Advances in Neural Information Processing Systems. 2017. P. 3146–3154.
  9. Chen T., Guestrin C. XGBoost: A Scalable Tree Boosting System //arXiv:1603.02754v3, 2016.
  10. Dorogush A., Prokhorenkova L., Gusev G. CatBoost: Unbiased Boosting with Categorical Features // arXiv:1706.09516v5, 2019.
  11. Имплементация библиотеки для подбора гиперпараметров Optuna на Python // GitHub. Optuna: webcite https://github.com/optuna/optuna (accessed: 07.02.2025).
  12. Имплементация библиотеки для подбора гиперпараметров HyperOpt на Python // GitHub. HyperOpt: webcite https://github.com/hyperopt/hyperopt (accessed: 07.02.2025).
  13. Fazulyanov D., Guseva A. Adaptive Recommendation System for Media Services: Analysis of User Interactions and Their Impact on Content Personalization // Technical sciences. 2024. № 5. P. 82–88.
  14. Hamed L., Abbar S., Haouari A. The Impact of Negative Preferences on a Recommendation Process // Intern. Conf. on Multimedia Computing and Systems (IEEE). Tangiers, Morocco, 2012. P. 675–680.
  15. Ma H., Xie R., Meng L., Feng F. Negative Sampling in Recommendation: A Survey and Future Directions // arXiv:2409.07237, 2024.
  16. Paudel B., Luck S., Bernstein A. Loss Aversion in Recommender Systems: Utilizing Negative User Preference to Improve Recommendation Quality // arXiv:1812.11422, 2018.
  17. He X., Liao L. Neural Collaborative Filtering // arXiv:1708.05031, 2017.
  18. Rendle S., Freudenthaler C. BPR: Bayesian Personalized Ranking from Implicit Feedback // arXiv:1205.2618, 2012.
  19. Weston J., Bengio S., Usunier N. Scaling Up To Large Vocabulary Image Annotation // Proc. 22nd Intern. Joint Conf. on Artificial Intelligence. Barcelona, Spain, 2011. P. 2764–2770.
  20. Lin T., Goyal P. Focal Loss for Dense Object Detection //arXiv:1708.02002, 2017.
  21. Wu Y., Xie R. DFGNN: Dual-frequency Graph Neural Network for Sign-aware Feedback // arXiv:2405.15280, 2024.
  22. Lin G., Gao C. Dual-interest Factorization-heads Attention for Sequential Recommendation // Proc. ACM Web Conf. Ostin, USA, 2017. P. 917–927.
  23. Vaswani A., Shazeer N., Parma N. Attention Is All You Need // arXiv:1706.03762, 2017.
  24. Kang W., McAuley J. Self-Attentive Sequential Recommendation // arXiv:1808.09781, 2018.
  25. Sun F., Liu J. BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformer // arXiv:1904.06690, 2019.
  26. Pereira Moreira G., Rabhi S. Transformers4Rec: Bridging the Gap between NLP and Sequential / Session-Based Recommendation // Proc. 15th ACM Conf. on Recommender Systems. Amsterdam, Netherlands, 2021. P. 143–153.
  27. Wang Y., Xun J. EAGER: Two-Stream Generative Recommender with Behavior-Semantic Collaboration // arXiv:2406.14017v1, 2024.
  28. Gao M., Zhang J. Recommender Systems Based on Generative Adversarial Networks: A Problem-Driven Perspective // arXiv:2003.02474, 2020.
  29. Liu Z., Ma Y. Contrastive Learning for Recommender System // arXiv:2101.01317, 2021.
  30. Ye H., Li X. On the Sweet Spot of Contrastive Views for Knowledge-enhanced Recommendation // arXiv:2309.13384, 2023.
  31. Serrano N. Bellogin A. Siamese Neural Networks in Recommendation // Neural Computing and Applications. 2023. V. 35. P. 13941–13953.
  32. Chen X. Yao L. Deep Reinforcement Learning in Recommender Systems: A Survey and New Perspectives // Knowledge-Based Systems. 2023. V. 264. № 110335.
  33. Ie E., Jain V. Reinforcement Learning for Slate-based Recommender Systems: A Tractable Decomposition and Practical Methodology // arXiv:1905.12767, 2019.
  34. Cena F., Console L., Vernero F. How to Deal with Negative Preferences in Recommender Systems: a Theoretical Framework // J. Intelligent Information Systems. 2023. V. 60. P. 23–47.
  35. Alzubaidi L., Bai J., Al-Sabaawi A. A Survey on Deep Learning Tools Dealing with Data Scarcity: Definitions, Challenges, Solutions, Tips, and Applications // J. Big Data. 2023. V. 10. № 46.
  36. Grinsztajn L., Oyallon E., Varoquaux G. Why Do Tree-Based Models Still Outperform Deep Learning on Tabular Data? // arXiv:2207.08815, 2022.
  37. Alzubaidi L., Zhang J., Humaidi A. Review of Deep Learning: Concepts, CNN Architectures, Challenges, Applications, Future Directions // J. Big Data. 2021. V. 8. № 53.
  38. Borisov V., Leemann T., Sebler K. Deep Neural Networks and Tabular Data: A Survey // IEEE Transactions on Neural Networks and Learning Systems. 2024. V. 35. № 6. P 7499–7519.
  39. Bentejac C., Csorgo A., Martinez-Munoz G. A Comparative Analysis of Gradient Boosting Algorithms // Artificial Intelligence Review. 2021. V. 54. P. 1937–1967.
  40. Sahour H., Gholami V., Torkaman J. Random Forest and Extreme Gradient Boosting Algorithms for Streamflow Modeling Using Vessel Features and Tree-Rings // Environmental Earth Sciences. 2021. V. 80. № 747.
  41. Demidova L., Sharshatov M., Shykhyev A. Methods for Solving the Class Imbalance Problem in Binary Classification Task // Information Technology and Data Standardization. 2024. № 3. P. 22–33.
  42. Wang Y., Halpern Y., Chan S. Learning from Negative User Feedback and Measuring Responsiveness for Sequential Recommenders // Proc. 17th ACM Conf. on Recommender System. Singapore, Singapore, 2023. P. 1049–1053.
  43. Paudel B., Luck S., Bernstein A. Loss Aversion in Recommender Systems: Utilizing Negative User Preference to Improve Recommendation Quality // arXiv:1812.11422, 2018.
  44. Wang X., Wu Y. An Improved HEAPSORT Algorithm with Nlogn – 0.788928n Comparisons in the Worst Case // J. Computer Science and Technology. 2007. V. 22. P. 898–903.
  45. Zhang X., Wang H., Liu Y. Retention Depolarization in Recommender System // Proc. ACM Web Conf. Singapore, Singapore. 2024. P. 1126–1137.
  46. Jadon A., Patil A. A Comprehensive Survey of Evaluation Techniques for Recommendation Systems // arXiv:2312.16015, 2024.
  47. Имплементация модели CatBoost на Python // GitHub. CatBoost: webcite https://github.com/catboost/catboost (accessed: 07.02.2025).

Supplementary files

Supplementary Files
Action
1. JATS XML
Download
2. Fig. 1. Scheme of the structure of generative adversarial networks.

Download (86KB)
Indexing metadata
3. Fig. 2. Experiment #1: Proportions of target events in samples.

Download (177KB)
Indexing metadata
4. Fig. 3. Experiment #1: sample sizes.

Download (274KB)
Indexing metadata
5. Fig. 4. Experiment #1: ROC-AUC values.

Download (283KB)
Indexing metadata
6. Fig. 5. Distribution of shares of object combinations.

Download (243KB)
Indexing metadata
7. Fig. 6. Experiment No. 1: accuracy of hitting the most probable combination.

Download (703KB)
Indexing metadata
8. Fig. 7. Experiment #2: Proportions of target events in the training set.

Download (264KB)
Indexing metadata
9. Fig. 8. Comparison of ROC-AUC for two experiments.

Download (384KB)
Indexing metadata

Copyright (c) 2025 Russian Academy of Sciences