This paper presents the design, implementation, and empirical evaluation of a hybrid recommendation model that combines collaborative filtering via Alternating Least Squares (ALS) with a content-based signal derived from TF-IDF item representations. The two signals are fused through an activity-adaptive user weighting α(u) that assigns higher weight to content when the user's training history is short and higher weight to collaborative evidence when the user has sufficient interactions. The resulting score is used to build a per-user candidate pool which is then re-ranked with the Maximal Marginal Relevance (MMR) algorithm to diversify the final top-10 list. Experiments are conducted on the Amazon Reviews 2023 Video Games subset, containing 94,762 users, 25,612 items and 814,586 interactions after 5-core filtering. Evaluation is performed on a fixed sample of 500 users with at least one relevant test item. With the empirically selected trade-off parameter λ = 0.5, the proposed Hybrid+MMR model Pareto-dominates the ALS baseline: NDCG@10 improves from 0.0240 to 0.0281 (+17%), HR@10 rises from 0.078 to 0.082 (+5%), and catalog coverage expands from 1.24% to 8.30%, a 6.7-fold increase. Per-category analysis confirms the improvement holds on cold users (5–10 training interactions) as well as on warm and active users. A lambda-sensitivity study reveals that the accuracy curve is concave with a single interior maximum, which justifies treating MMR's λ as a tunable hyper-parameter rather than fixing it from domain intuition. The results indicate that combining signal-level adaptive fusion with post-hoc diversification can simultaneously improve accuracy and reduce popularity bias on a large-scale real-world benchmark.

IBRAGIM D.A.

Master's student, Astana IT University, Astana, Kazakhstan.

E-mail: 242888@astanait.edu.kz, https://orcid.org/0009-0009-4830-1957

ZHUMADILLAYEVA A.K.

Candidate of technical sciences, associate professor, school of software engineering, Astana IT University, Astana, Kazakhstan.

E-mail: Ainur.Zhumadillayeva@astanait.edu.kz, https://orcid.org/0000-0003-1042-0415

1. Roy D., Dutta M. A systematic review and research perspective on recommender systems. Journal of Big Data. 2022. Vol. 9, No. 1. Art. 59. DOI: 10.1186/s40537-022-00592-5. DOI: https://doi.org/10.1186/s40537-022-00592-5
2. Duricic T., Kowald D., Lacic E., Lex E. Beyond-accuracy: a review on diversity, serendipity, and fairness in recommender systems based on graph neural networks. Frontiers in Big Data. 2023. Vol. 6. Art. 1251072. DOI: 10.3389/fdata.2023.1251072. DOI: https://doi.org/10.3389/fdata.2023.1251072
3. Raza S., Rahman M., Kamawal S., Toroghi A., Raval A., Navah F., Kazemeini A. A comprehensive review of recommender systems: transitioning from theory to practice. arXiv preprint. 2024. arXiv:2407.13699. DOI: 10.48550/arXiv.2407.13699.
4. Klimashevskaia A., Jannach D., Elahi M., Trattner C. A survey on popularity bias in recommender systems. User Modeling and User-Adapted Interaction. 2024. Vol. 34, No. 5. P. 1777–1834. DOI: 10.1007/s11257-024-09406-0. DOI: https://doi.org/10.1007/s11257-024-09406-0
5. Sami A., El Adrousy W., Sarhan S., Elmougy S. A deep learning based hybrid recommendation model for internet users. Scientific Reports. 2024. Vol. 14. Art. 28781. DOI: 10.1038/s41598-024-79011-z. DOI: https://doi.org/10.1038/s41598-024-79011-z
6. Bodduluri K.C., Palaniappan S., Kumar P.M.R., Selvarajan S. Hybrid recommender systems: survey and taxonomy. ACM Computing Surveys. 2024. Vol. 56, No. 10. P. 1–42. DOI: 10.1145/3652301. DOI: https://doi.org/10.1145/3661821
7. Liu Y., Yang J., Sun Y., Xu K., Chua T.-S. A review of deep learning and large language models for cold start problem in recommender systems. WIREs Data Mining and Knowledge Discovery. 2026. Vol. 16, No. 1. Art. e70068. DOI: 10.1002/widm.70068. DOI: https://doi.org/10.1002/widm.70068
8. Chen L., Wu L., Hong R., Zhang K., Wang M. Revisiting recommender systems: an investigative survey. Neural Computing and Applications. 2025. Vol. 37, No. 3. P. 1245–1278. DOI: 10.1007/s00521-024-10828-5. DOI: https://doi.org/10.1007/s00521-024-10828-5
9. Carbonell J., Goldstein J. The use of MMR, diversity-based reranking for reordering documents and producing summaries. Proceedings of the 21st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR '98). New York: ACM, 1998. P. 335–336. DOI: 10.1145/290941.291025. DOI: https://doi.org/10.1145/290941.291025
10. Kaminskas M., Bridge D. Diversity, serendipity, novelty, and coverage: a survey and empirical analysis of beyond-accuracy objectives in recommender systems. ACM Transactions on Interactive Intelligent Systems. 2017. Vol. 7, No. 1. Art. 2. DOI: 10.1145/2926720. DOI: https://doi.org/10.1145/2926720
11. Farzaneh M., Verma A. Enhancing RAG with Maximum Marginal Relevance (MMR) in Azure AI Search [Electronic resource]. 2024. URL: https://farzzy.hashnode.dev/enhancing-rag-with-maximum-marginal-relevance-mmr-in-azure-ai-search (accessed: 15.04.2026).
12. Hou Y., Li J., He Z., Yan A., Chen X., McAuley J. Bridging language and items for retrieval and recommendation. arXiv preprint. 2024. arXiv:2403.03952. DOI: 10.48550/arXiv.2403.03952.
13. Hu Y., Koren Y., Volinsky C. Collaborative filtering for implicit feedback datasets. Proceedings of the 8th IEEE International Conference on Data Mining (ICDM 2008). Pisa: IEEE, 2008. P. 263–272. DOI: 10.1109/ICDM.2008.22. DOI: https://doi.org/10.1109/ICDM.2008.22
14. Bhattacharya S., Ghosh R. SMMR: sampling-based MMR reranking for faster, more diverse, and balanced recommendations and retrieval. Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR '25). Padua: ACM, 2025. P. 2512–2521. DOI: 10.1145/3726302.3730250. DOI: https://doi.org/10.1145/3726302.3730250
15. Chen Z., Gan W., Wu J., Hu K., Lin H. Data scarcity in recommendation systems: a survey. ACM Transactions on Recommender Systems. 2025. Vol. 3, No. 3. P. 1–31. DOI: 10.1145/3686112. DOI: https://doi.org/10.1145/3639063
16. Jangid M., Kumar R. Deep learning approaches to address cold start and long tail challenges in recommendation systems: a systematic review. Multimedia Tools and Applications. 2025. Vol. 84, No. 5. P. 2293–2325. DOI: 10.1007/s11042-024-19871-9. DOI: https://doi.org/10.1007/s11042-024-20262-3
17. López-Ávila A., Du J. A survey on large language models in multimodal recommender systems. arXiv preprint. 2025. arXiv:2505.09777. DOI: 10.48550/arXiv.2505.09777.
18. Zhao Y., Wang Y., Liu Y., Cheng X., Aggarwal C., Derr T. Fairness and diversity in recommender systems: a survey. ACM Transactions on Intelligent Systems and Technology. 2024. Vol. 15, No. 5. Art. 97. DOI: 10.1145/3664928. DOI: https://doi.org/10.1145/3664928

recommender systems, hybrid recommendation, collaborative filtering, content-based filtering, MMR diversification, cold-start problem, catalog coverage