Abstract
Systems and methods can quantify user journey progression in digital advertising analytics, as some static, rule-based models may not capture the non-linear nature of certain user behaviors. The disclosed technology can utilize a data-driven framework with a sequential prediction model, such as an n-gram model, to analyze user interaction data. Raw user actions can be ingested and mapped to discrete symbols, or grams, which can then be formed into chronological sequences representing user journeys. The model can process these advertiser-specific sequences to compute a transition probability table that quantifies the likelihood of a user's next action based on their recent activity. This approach can provide a quantitative method for measuring user progression, yielding metrics, for example a predicted conversion probability score, that can be used for campaign analysis, audience segmentation, and automated bidding.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Wang, Lijun; Zhang, Qing; Vaughan, Thomas; Fu, Kangxing; Chung, Shuo-Heng; Skvortsov, Evgeny; Nair, Harikesh; Xu, Shawn; Wang, Jianqiao; Wang, Jing; Huyi, Charles; Zhang, Junyi; Quintero, David; Yam, Jeff; Root, Jiro; and Geng, Tong, "Quantifying User Journey Progression Using a Language Model", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/9186