Ad Network Content Classification Registry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current ad distribution systems lack a means to easily prevent advertisements from being served to certain types of mobile applications, due to server policies, advertiser criteria, and legal restrictions, leading to inappropriate content being displayed.
Innovation Solution
An ad network system that obtains and analyzes content information from mobile device applications to determine appropriateness for advertisement serving, using statistical machine learning techniques to classify content and update a registry to prevent ads from being served to applications with inappropriate content, such as adult content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If ads are distributed to all requesting apps without restriction, then ad network productivity is improved, but inappropriate content may be displayed violating policies and legal restrictions
Solution Approach 1:
The system performs preliminary classification of mobile device applications by obtaining content information from app servers and analyzing it before ad distribution. The ad server maintains a registry that stores classification results indicating whether each app is appropriate for ad serving, allowing the system to quickly determine ad eligibility without re-analyzing content each time an ad request is received.
Solution Approach 2:
The patent introduces an intermediary classification mechanism between the ad server and requesting apps. The ad server acts as a mediator that checks the registry to determine whether to serve ads to a particular app, preventing inappropriate content display while maintaining efficient ad distribution to legitimate apps.
2Reliability
If content classification and ad restriction mechanisms are implemented, then content appropriateness is improved, but system complexity increases
Solution Approach 1:
The patent extracts the content classification function from the ad serving process itself, separating it into a distinct preliminary step. The ad server obtains content information from app servers and performs classification independently, storing results in a registry. This separation allows the core ad serving mechanism to remain simple while adding reliability through the extracted classification function.
Solution Approach 2:
The system creates a simplified copy of content classification information by storing classification results in a registry rather than re-analyzing full app content each time. The registry contains essential appropriateness indicators that can be quickly referenced, reducing the complexity of real-time ad serving decisions while maintaining reliable content appropriateness verification.
3Measurement precision
If app content information is obtained and analyzed for each ad request, then content classification accuracy is improved, but processing time increases
Solution Approach 1:
The system performs content classification in advance and stores the results in a registry before actual ad serving occurs. When an ad request is received, the ad server simply checks the pre-computed classification results in the registry rather than analyzing app content in real-time, maintaining high classification accuracy while minimizing response time.
Solution Approach 2:
The system performs content classification once comprehensively to obtain accurate results, then uses these partial results (classification indicators) for subsequent ad serving decisions. This avoids repeatedly performing full content analysis for each ad request, significantly reducing processing time while maintaining classification accuracy through the stored registry information.
Data Source
AI summary
Application content information is obtained from a mobile device application server. An ad server log includes at least application identifiers for applications that have requested advertisements from an ad network. If identifying information included in the application content information is the same as an application identifier found in the ad server log, the application content information is analyzed to determine whether the corresponding mobile device application is appropriate to be served an advertisement, based on predetermined content criteria. An application registry includes a database of mobile device applications that are registered in the ad network. The application registry record for a mobile device application is updated to indicate the determined appropriateness of the mobile device application corresponding to the analyzed application content information.