Journal of Soft Computing and Information Technology

Journal of Soft Computing and Information Technology

Improving the Performance of The Recommender System by Solving the Cold Start Problem Using Frequent Pattern Extraction Based on Fuzzy Clustering

Document Type : Persian Original Article

Authors
Samaneh Zirak
Mehdi Salkhordeh Haghighi
Faculty of Computer Engineering and information technology, Sadjad University, Mashhad, Iran.
10.22034/jscit.2026.559654.2190
Abstract
The cold-start problem remains one of the major challenges in recommender system, as the lack of sufficient information about new users or items significantly reduces recommendation accuracy and reliability. This paper proposes a hybrid recommendation framework that combines Fuzzy C-Means (FCM) clustering, DBSCAN, and FP-Growth to alleviate the effects of the cold-start problem. After preprocessing the MovieLens datasets, DBSCAN is employed to determine the optimal number of user clusters, and FCM is subsequently applied to group users according to their behavioral similarities. Frequent behavioral patterns are then extracted from each cluster using the FP-Growth algorithm and utilized to generate personalized recommendations for users with similar preferences. The proposed approach is evaluated on the MovieLens 100K and MovieLens 1M datasets using Precision, Recall, and F-measure metrics. Experimental results demonstrate that the proposed method achieves Precision, Recall, and F-measure values of 0.78, 0.76, and 0.77 on the MovieLens 100K dataset, and 0.82, 0.77, and 0.795 on the MovieLens 1M dataset, respectively. To further address the new-user cold-start problem, three initialization strategies, including content-based, demographic-based, and questionnaire-based approaches, are investigated. Among them, the questionnaire-based approach yields the best performance, achieving a Precision value of 0.654. Comparative analyses indicate that the proposed framework effectively improves recommendation quality and reduces the impact of the cold-start problem compared with existing approaches.
Keywords
Cold start problem
recommender systems
fuzzy clustering
Fuzzy C-Means (FCM), DBSCAN, FP-Growth, collaborative filtering, frequent pattern mining
[1] R. D. Gottapu and L. V. Sriram Monangi, "Point-Of-Interest Recommender System for Social Groups," Procedia Comput. Sci., vol. 114, pp. 159–164, 2017.
[2] J. Gao, T. Zhang, and C. Xu, "A unified personalized video recommendation via dynamic recurrent neural networks," in Proc. 2017 ACM Multimed. Conf., 2017, pp. 127–135.
[3] S. Lakshmi and Dr. T. Adi Lakshmi, "Recommendation Systems: Issues and challenges," Int. J. Comput. Sci. Inform. Technol. (IJCSIT), vol. 5, no. 4, pp. 5771–5772, 2014.
[4] F. Fadhel Aljunid and D. H. Manjaiah, "A Survey on Recommendation Systems for Social Media Using big data Analytics," Int. J. Latest Trends Eng. Technol., Special Issue (SACAIM 2017), pp. 48–58.
[5] J. Ma, G. Li, M. Zhong, X. Zhao, L. Zhu, and X. Li, "LGA: latent genre aware micro-video recommendation on social media," Multimed. Tools Appl., vol. 77, no. 3, pp. 2991–3008, 2018.
[6] P. Priyanga and A. R. N. Banu Kamal, "Methods of Mining the Data from Big Data and Social Networks Based on Recommender System," Int. J. Adv. Netw. Appl. (IJANA), vol. 8, no. 5, pp. 55–60, 2017.
[7] M. Wasid and R. Ali, "An improved recommender system based on multi-criteria clustering approach," Procedia Comput. Sci., vol. 131, pp. 93–101, 2018.
[8] M. H. Mohamed, M. H. Khafagy, and M. H. Ibrahim, "Recommender Systems Challenges and Solutions Survey," in Proc. 2019 Int. Conf. Innov. Trends Comput. Eng. ITCE, Feb. 2019, pp. 149–155.
[9] M. Tsagkias, T. H. King, S. Kallumadi, V. Murdock, and M. de Rijke, "Challenges and research opportunities in eCommerce search and recommendations," ACM SIGIR Forum, vol. 54, no. 1, pp. 1–23, Feb. 2021.
[10] F. Merabet and D. Benmerzoug, "QoS prediction for service selection and recommendation with a deep latent features autoencoder," Comput. Sci. Inf. Syst., vol. 19, no. 2, pp. 709–733, 2022.
[11] P. M. Alamdari, N. J. Navimipour, M. Hosseinzadeh, A. A. Safaei, and A. Darwesh, "A systematic study on the recommender systems in the e-commerce," IEEE Access, vol. 8, pp. 115694–115716, 2020.
[12] A. Pawlicka, M. Pawlicki, R. Kozik, and R. S. Choraś, "A systematic review of recommender systems and their applications in cybersecurity," Sensors, vol. 21, no. 15, p. 5248, Aug. 2021.
[13] Z. Batmaz, A. Yurekli, A. Bilge, and C. Kaleli, "A review on deep learning for recommender systems: Challenges and remedies," Artif. Intell. Rev., vol. 52, no. 1, pp. 1–37, Jun. 2019.
[14] E. Kannout, M. Grodzki, and M. Grzegorowski, "Utilizing frequent pattern mining for solving cold-start problem in recommender systems," in Proc. 17th Conf. Comput. Sci. Intell. Syst. (FedCSIS), Sofia, Bulgaria, Sep. 2022, pp. 217–226.
[15] E. Kannout, M. Grodzki, M. Grodzki, and H. Nguyn, "Clustering-Based Frequent Pattern Mining Framework for Solving Cold-Start Problem in Recommender Systems," IEEE Access, 2024, pp. 13678–13698.
[16] M. Grodzki and M. Grzegorowski, "Utilizing Frequent Pattern Mining for Solving Cold-Start Problem in Recommender Systems," in Proc. 17th Conf. Comput. Sci. Intell. Syst., 2022, pp. 217–226.
[17] M. L. Lalu Zazuli Azhar Mardedi, Kartarina, & Moch. Syahrir, "Analisis Perbandingan Algoritma Fp-Growth dan Tpq-Apriori dalam Menentukan Rule Based Terbaik untuk Sistem Rekomendasi Produk, " Jurnal Sistem Informasi dan Telematika, Vol. 14, No. 2, pp. – –, Jul. 2024, doi:10.35200/ex.v14i2.112.
[18] E. Kannout, M. Grodzki, and M. Grzegorowski, "Towards addressing item cold-start problem in collaborative filtering by embedding agglomerative clustering and FP-Growth into the recommendation system," Comput. Sci. Inf. Syst., vol. 20, no. 4, pp. 1343–1366, 2022.
[19] N. Musa Al-Rezaei Golian, "Designing a hybrid recommender system to solve the cold start problem using clustering and chaotic particle swarm algorithm," Sci. J. Soft Comput., pp. 50–61,2024.
[20] M. Kawai, H. Sato, and T. Shiohama, "Topic model-based recommender systems and their applications to cold-start problems," Expert Syst. Appl., vol. 202, p. 117129, Sep. 2022.
[21] M. Pulis and J. Bajada, "Siamese neural networks for content-based cold-start music recommendation," in Proc. 15th ACM Conf. Recommender Syst., New York, NY, USA: ACM, 2021, pp. 719–723.
[22] J. Hu, T. An, Z. Yu, J. Du, and Y. Luo, "Contrastive Learning for Cold Start Recommendation with Adaptive Feature Fusion,"  in Proc. 2025 5th Int. Conf. Consumer Electronics and Computer Engineering (ICCECE), Dongguan, China, Feb. 28–Mar. 2, 2025, doi:10.1109/ICCECE65250.2025.10985689.
[23] X. Guo, T. Zhang, F. Wang, X. Wang, X. Zhang, X. Liu, and Z. Cui, "MMHCL: Multi-Modal Hypergraph Contrastive Learning for Recommendation,” arXiv preprint arXiv:2504.16576, 2025.
[24] C. He, Y. Liu, Q. Li, C. Hong, W. Zhong, and X. Yao, "M²VAE: Multi-Modal Multi-View Variational Autoencoder for Cold-start Item Recommendation," arXiv preprint arXiv:2508.00452, 2025.
Journal of Soft Computing and Information Technology
Volume 15, Issue 3
Autumn - October
Autumn 2026
Pages 21-35