Advertisement Component Review Caching for Approval Speed
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual review of advertisements in online systems is time-consuming and inconsistent, leading to increased review times and rejection rates, as well as uncertainty for advertisers regarding approval likelihood, which affects revenue generation.
Innovation Solution
Divide advertisements into components and store analyzed components in a database, allowing the system to reuse previous reviews for matching components, reducing the number of reviews needed and improving consistency through efficient computational resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual review of advertisements is performed, then compliance with policies can be ensured, but review time increases and productivity decreases
Solution Approach 1:
The patent segments an advertisement into multiple components (e.g., title, description, image, link) and reviews each component separately. This allows the system to efficiently process and cache results for individual components, reducing overall review time while maintaining compliance accuracy through systematic evaluation of each segment.
Solution Approach 2:
The patent performs preliminary review and caching of advertisement components before final advertisement submission. By pre-reviewing components that may be reused across multiple advertisements and caching their compliance results, the system reduces the time required for subsequent reviews while ensuring policy compliance is established in advance.
2Reliability
If manual review of advertisements is performed, then policy compliance can be checked, but inconsistencies arise due to different reviewer views
Solution Approach 1:
The patent implements an automated review system that evaluates advertisement components against stored policy rules without human intervention. The system serves itself by automatically comparing components to compliance criteria, eliminating subjective variations between different reviewers and ensuring consistent application of policies across all advertisements.
Solution Approach 2:
The patent establishes a feedback mechanism where review results are cached and used to inform subsequent review decisions. The system continuously learns from previous review outcomes and applies this knowledge to maintain consistent compliance checking, ensuring that the same policy interpretations are applied uniformly across different advertisements and reviewers.
3Device complexity
If advertisements are reviewed without component analysis, then review process is simpler, but review time increases due to re-evaluating same components multiple times
Solution Approach 1:
The patent divides advertisements into discrete components that can be independently analyzed and cached. By segmenting the review process at the component level rather than reviewing entire advertisements repeatedly, the system eliminates redundant evaluation of the same elements while maintaining manageable complexity through organized component handling.
Solution Approach 2:
The patent implements a caching mechanism that stores review results for advertisement components. When the same or similar components appear in new advertisements, the system recovers previously cached results instead of re-evaluating them, discarding redundant computation while preserving valuable review insights for future use.
Data Source
AI summary
As components of an advertisement are received by an online system, the online system determines whether the received components satisfy one or more policies of the online system. If a component does not satisfy a policy of the online system, the user providing the components is notified when the component is provided, allowing the user to modify the component. Additionally, stored data describing previously analyzed components are compared to received components, and the stored data is used if the received component matches a previously analyzed component. This process expedites approval of advertisements by the online system.