User Grouping Optimization for Ad Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for optimizing online advertisement delivery often limit the user base by targeting only users with high conversion rates, leading to compromised overall user exposure and unfairness in advertisement distribution.
Innovation Solution
A system and method that divide users into user buckets based on conversion scores, using a trained prediction model to maximize a total conversion score through an optimization model, assigning user buckets to groups to achieve a balanced and optimal user grouping strategy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If users are divided into a few groups according to conversion rates with only top users receiving advertisements, then short-term conversion benefit is improved, but overall user exposure and fairness are compromised
Solution Approach 1:
The patent divides users into multiple buckets based on conversion rates and further segments these buckets into different user groups. Instead of treating all users uniformly or only targeting top users, the system creates fine-grained segments that allow differentiated advertisement delivery strategies, thereby expanding the effective user base while maintaining conversion efficiency.
Solution Approach 2:
The patent changes the grouping parameter from simple conversion rate ranking to a multi-dimensional approach considering bucket indices, group assignments, and optimization objectives. By adjusting these parameters, the system can balance between targeting high-conversion users and expanding overall user exposure, resolving the contradiction between conversion rate and user base quantity.
2Device complexity
If users are divided into a few groups according to conversion rates, then advertisement delivery is simplified, but user management granularity and accuracy are reduced
Solution Approach 1:
The patent implements multi-level segmentation: first dividing users into buckets based on conversion rates, then assigning these buckets to different user groups. This hierarchical segmentation achieves fine-grained user management without proportionally increasing system complexity, as the bucketing step provides a structured foundation for subsequent group assignments.
Solution Approach 2:
The patent introduces dynamic optimization where the assignment of user buckets to groups is not fixed but determined through optimization models that consider various objectives. This dynamic approach allows the system to adapt grouping strategies based on specific advertisement campaigns and performance goals, improving management accuracy without permanent complexity increases.
3Ease of manufacture
If traditional user grouping methods are used, then implementation is straightforward, but fairness and overall effectiveness in online advertising are limited
Solution Approach 1:
The patent introduces optimization models and prediction models as intermediaries between traditional grouping methods and advertisement delivery. These models act as mediators that translate simple user segmentation into optimized group assignments, maintaining implementation feasibility while significantly improving overall advertising effectiveness through data-driven decision-making.
Solution Approach 2:
The patent incorporates feedback mechanisms where user responses to advertisements are continuously monitored and used to refine bucket assignments and group configurations. This feedback loop enables the system to learn from past performance and continuously improve effectiveness while maintaining a structured implementation approach that builds on traditional grouping concepts.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for grouping users are provided. One embodiment of the methods includes: dividing a plurality of users targeted by the an advertisement candidate into a plurality of user buckets, wherein each of the plurality of user buckets is associated with a first conversion score; obtaining a trained prediction model corresponding to the advertisement, wherein the trained prediction model is able to predict a conversion score based at least on the first conversion score associated with a user bucket and a second conversion score associated with a group of user buckets comprising the user bucket; and constructing an optimization model using the trained prediction model, wherein an objective function of the optimization problem is to maximize a total conversion score with a grouping strategy determined by solving the optimization problem.