SeqUDA-Rec: Sequential User Behavior Enhanced Recommendation via Global Unsupervised Data Augmentation for Personalized Content Marketing
DOI:
https://doi.org/10.71222/1r79aa27Keywords:
sequential recommendation, user behavior analysis, graph contrastive learning, GAN-based augmentation, personalized content marketingAbstract
Personalized content marketing has become a central strategy for digital platforms seeking to deliver highly relevant advertisements and recommendations. However, conventional recommendation systems often rely heavily on sparse supervised signals derived from explicit user feedback and are easily affected by noisy or unintentional interactions, resulting in unstable predictions. To overcome these limitations, we propose SeqUDA-Rec, a unified deep learning framework that incorporates sequential user behavior modeling with global unsupervised data augmentation. The framework begins by constructing a Global User-Item Interaction Graph (GUIG) from all historical behavior sequences, enabling the extraction of both local item transitions and global cross-user relational structures. A graph contrastive learning module is introduced to learn noise-resistant representations by maximizing agreement across multiple graph views. Meanwhile, a Transformer-based sequential encoder captures users' evolving interests and long-term dependencies within interaction trajectories. To further mitigate the challenges of limited labeled data and behavior sparsity, SeqUDA-Rec integrates a GAN-based augmentation module, which generates realistic synthetic sub-sequences to enrich training diversity and improve model generalization. We evaluate SeqUDA-Rec on two large-scale advertising datasets-Amazon Ads and TikTok Ad Clicks-covering both stable e-commerce environments and fast-changing short-video scenarios. Experimental results show that our model consistently outperforms strong baselines including SASRec, BERT4Rec, and GCL4SR, achieving 6.7% improvement in NDCG@10 and 11.3% improvement in HR@10. These results demonstrate that SeqUDA-Rec effectively enhances recommendation robustness, alleviates noise sensitivity, and provides a powerful solution for real-world personalized content marketing.
References
1. H. Matharu, Z. Pasha, M. Aarif, L. Natrayan, S. Kaliappan, and I. I. Raj, "Developing an AI-Driven Personalization Engine for Real-Time Content Marketing in E-commerce Platforms," In 2024 15th International Conference on Computing Communication and Networking Technologies (ICCCNT), June, 2024, pp. 1-6. doi: 10.1109/icccnt61001.2024.10725400
2. A. S. Tewari, and A. G. Barman, "Sequencing of items in personalized recommendations using multiple recommendation techniques," Expert Systems with Applications, vol. 97, pp. 70-82, 2018. doi: 10.1016/j.eswa.2017.12.019
3. N. Yuill, and Y. Rogers, "Mechanisms for collaboration: A design and evaluation framework for multi-user interfaces," ACM Transactions on Computer-Human Interaction (TOCHI), vol. 19, no. 1, pp. 1-25, 2012.
4. A. M. de Sousa, "Analyzing and modeling user curiosity in online information services," 2024.
5. Z. Fatima Ezzahra, A. Sana, Q. Sara, and R. Said, "Multi-objective reinforcement learning for recommender systems: a comprehensive survey of methods, challenges, and future directions," International Journal of Multimedia Information Retrieval, vol. 14, no. 4, pp. 1-35, 2025. doi: 10.1007/s13735-025-00383-7
6. M. Jing, Y. Zhu, T. Zang, and K. Wang, "Contrastive self-supervised learning in recommender systems: A survey," ACM Transactions on Information Systems, vol. 42, no. 2, pp. 1-39, 2023. doi: 10.1145/3627158
7. X. Ren, W. Wei, L. Xia, and C. Huang, "A comprehensive survey on self-supervised learning for recommendation," ACM Computing Surveys, vol. 58, no. 1, pp. 1-38, 2025. doi: 10.1145/3746280
8. F. T. Abdul Hussien, A. M. S. Rahma, and H. B. Abdulwahab, "An e-commerce recommendation system based on dynamic analysis of customer behavior," Sustainability, vol. 13, no. 19, p. 10786, 2021. doi: 10.3390/su131910786
9. R. V. Karthik, and S. Ganapathy, "A fuzzy recommendation system for predicting the customers interests using sentiment analysis and ontology in e-commerce," Applied Soft Computing, vol. 108, p. 107396, 2021.
10. K. Xu, H. Zhou, H. Zheng, M. Zhu, and Q. Xin, "Intelligent classification and personalized recommendation of e-commerce products based on machine learning," arXiv preprint arXiv:2403.19345, 2024. doi: 10.54254/2755-2721/64/20241365
11. R. Luo, X. Chen, and Z. Ding, "SeqUDA-Rec: Sequential user behavior enhanced recommendation via global unsupervised data augmentation for personalized content marketing," arXiv preprint arXiv:2509.17361, 2025.
12. A. L. Karn, R. K. Karna, B. R. Kondamudi, G. Bagale, D. A. Pustokhin, I. V. Pustokhina, and S. Sengan, "RETRACTED ARTICLE: Customer centric hybrid recommendation system for E-Commerce applications by integrating hybrid sentiment analysis," Electronic commerce research, vol. 23, no. 1, pp. 279-314, 2023. doi: 10.1007/s10660-022-09630-z
