New User Event Prediction Through the Lens of Causal Inference

Henry Yuchi, Shixiang Zhu, Li Dong, Yigit M. Arisoy, Matthew C. Spencer
Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, PMLR 258:2953-2961, 2025.

Abstract

Modeling and analysis for event series generated by users of heterogeneous behavioral patterns are closely involved in our daily lives, including credit card fraud detection, online platform user recommendation, and social network analysis. The most commonly adopted approach to this task is to assign users to behavior-based categories and analyze each of them separately. However, this requires extensive data to fully understand the user behavior, presenting challenges in modeling newcomers without significant historical knowledge. In this work, we propose a novel discrete event prediction framework for new users with limited history, without needing to know the user’s category. We treat the user event history as the "treatment" for future events and the user category as the key confounder. Thus, the prediction problem can be framed as counterfactual outcome estimation, where each event is re-weighted by its inverse propensity score. We demonstrate the improved performance of the proposed framework with a numerical simulation study and two real-world applications, including Netflix rating prediction and seller contact prediction for customer support at Amazon.

Cite this Paper

BibTeX
@InProceedings{pmlr-v258-yuchi25a, title = {New User Event Prediction Through the Lens of Causal Inference}, author = {Yuchi, Henry and Zhu, Shixiang and Dong, Li and Arisoy, Yigit M. and Spencer, Matthew C.}, booktitle = {Proceedings of The 28th International Conference on Artificial Intelligence and Statistics}, pages = {2953--2961}, year = {2025}, editor = {Li, Yingzhen and Mandt, Stephan and Agrawal, Shipra and Khan, Emtiyaz}, volume = {258}, series = {Proceedings of Machine Learning Research}, month = {03--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v258/main/assets/yuchi25a/yuchi25a.pdf}, url = {https://proceedings.mlr.press/v258/yuchi25a.html}, abstract = {Modeling and analysis for event series generated by users of heterogeneous behavioral patterns are closely involved in our daily lives, including credit card fraud detection, online platform user recommendation, and social network analysis. The most commonly adopted approach to this task is to assign users to behavior-based categories and analyze each of them separately. However, this requires extensive data to fully understand the user behavior, presenting challenges in modeling newcomers without significant historical knowledge. In this work, we propose a novel discrete event prediction framework for new users with limited history, without needing to know the user’s category. We treat the user event history as the "treatment" for future events and the user category as the key confounder. Thus, the prediction problem can be framed as counterfactual outcome estimation, where each event is re-weighted by its inverse propensity score. We demonstrate the improved performance of the proposed framework with a numerical simulation study and two real-world applications, including Netflix rating prediction and seller contact prediction for customer support at Amazon.} }
Endnote
%0 Conference Paper %T New User Event Prediction Through the Lens of Causal Inference %A Henry Yuchi %A Shixiang Zhu %A Li Dong %A Yigit M. Arisoy %A Matthew C. Spencer %B Proceedings of The 28th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2025 %E Yingzhen Li %E Stephan Mandt %E Shipra Agrawal %E Emtiyaz Khan %F pmlr-v258-yuchi25a %I PMLR %P 2953--2961 %U https://proceedings.mlr.press/v258/yuchi25a.html %V 258 %X Modeling and analysis for event series generated by users of heterogeneous behavioral patterns are closely involved in our daily lives, including credit card fraud detection, online platform user recommendation, and social network analysis. The most commonly adopted approach to this task is to assign users to behavior-based categories and analyze each of them separately. However, this requires extensive data to fully understand the user behavior, presenting challenges in modeling newcomers without significant historical knowledge. In this work, we propose a novel discrete event prediction framework for new users with limited history, without needing to know the user’s category. We treat the user event history as the "treatment" for future events and the user category as the key confounder. Thus, the prediction problem can be framed as counterfactual outcome estimation, where each event is re-weighted by its inverse propensity score. We demonstrate the improved performance of the proposed framework with a numerical simulation study and two real-world applications, including Netflix rating prediction and seller contact prediction for customer support at Amazon.
APA
Yuchi, H., Zhu, S., Dong, L., Arisoy, Y.M. & Spencer, M.C.. (2025). New User Event Prediction Through the Lens of Causal Inference. Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 258:2953-2961 Available from https://proceedings.mlr.press/v258/yuchi25a.html.

Related Material