AI-Driven Intelligent Cold Start Ad Click Prediction System for Advertisement Recommendation
Keywords:
Click-through rate (CTR), Cold-start problem, Digital advertising, Machine learning, Predictive analytics, Random forest, User engagement predictionAbstract
Ad serving technologies are highly dependent on the accurate prediction of the click-through rate (CTR). Still, the vast majority of existing prediction models utilise data on past users’ behaviour; thus, they do not provide any predictions for new or anonymous users. Consequently, one encounters the cold-start problem, leading to inadequate ad placement and ineffective utilisation of advertising space. To mitigate this issue, this project suggests implementing a machine learning approach to predict user engagement independent of their past behaviour. Demographic attributes, such as age, gender, and country, along with other factors, such as device and time of interaction, will be incorporated into the model. The core of the model is the random forest algorithm due to its efficiency and accuracy in dealing with non-linear problems. The developed solution will produce probability-based predictions for further advertisement selection. Also, the implementation of analytics functionality will allow gaining insights, such as overall CTR and prediction probability values.
References
H. B. McMahan et al., “Ad click prediction,” Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD ’13, 2013, pp. 1222–1230.
S. Rendle, “Factorization machines,” 2010 IEEE International Conference on Data Mining, Sydney, NSW, Australia, 2010, pp. 995–1000.
H.-T. Cheng et al., “Wide & deep learning for recommender systems,” arXiv:1606.07792 [cs, stat], Jun. 2016.
H. Guo, R. Tang, Y. Ye, Z. Li, and X. He, “DeepFM: A factorization-machine based neural network for CTR prediction,” arXiv:1703.04247v1, Mar. 2017.
Y. Juan, Y. Zhuang, W.-S. Chin, and C.-J. Lin, “Field-aware factorization machines for CTR prediction,” Proceedings of the 10th ACM Conference on Recommender Systems - RecSys ’16, 2016.
Y. Tagami, S. Ono, K. Yamamoto, K. Tsukamoto, and A. Tajima, “CTR prediction for contextual advertising,” Proceedings of the Seventh International Workshop on Data Mining for Online Advertising, pp. 1–8, Aug. 2013.
Y. Yang and P. Zhai, “Click-through rate prediction in online advertising: A literature review,” Information Processing & Management, vol. 59, no. 2, p. 102853, Mar. 2022.
W. Zhang, J. Qin, W. Guo, R. Tang, and X. He, “Deep learning for click-through rate estimation,” arXiv:2104.10584 [cs], Apr. 2021.
J. Zhu, J. Liu, S. Yang, Q. Zhang, and X. He, “Open benchmarking for click-through rate prediction,” arXiv.org, 2020.
J. Bai et al., “A comprehensive survey on advertising click-through rate prediction algorithm,” The Knowledge Engineering Review, vol. 40, Jan. 2025.
