Ad Selection System Using Content-Based Intermediary
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for selecting and displaying advertisements over the internet are limited in targeting relevance and effectiveness, as they either restrict audience reach or invade user privacy, and lack control over ad selection once the user navigates away from the initial web page.
Innovation Solution
A system and method for selecting and displaying video advertisements based on relevance, bid price, past performance, or a combination thereof, using a database and ad selection application that receives requests from user computers, incorporates user demographic data, and orders ads according to criteria such as location, format, and relevancy scores, allowing for real-time selection and display within web pages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If advertisements are selected based on user demographic information collected through questionnaires or cookies, then ad targeting relevance is improved, but user privacy is invaded
Solution Approach 1:
The patent introduces a content-based intermediary system that analyzes web page content, metadata, and contextual information as a mediator between advertisers and users. This intermediary approach enables ad selection based on page relevance without directly collecting or processing sensitive user demographic data, thus maintaining privacy while achieving targeting relevance.
Solution Approach 2:
The patent extracts and utilizes publicly available content information from web pages, metadata, and contextual data as a substitute for invasive user demographic collection. By taking out and leveraging content-based signals, the system achieves ad targeting without requiring personal user information.
2Measurement precision
If advertisements are displayed only on specific web pages through contractual agreements, then ad-relevant content matching is improved, but audience reach is limited
Solution Approach 1:
The patent creates a universal ad selection system that can operate across multiple web pages and platforms by analyzing content and contextual signals. This multi-functional approach allows the same ad selection mechanism to serve diverse audiences across different web pages while maintaining content-relevance, thereby expanding audience reach without sacrificing matching precision.
Solution Approach 2:
The patent implements dynamic ad selection that adapts to different web page contexts, user behaviors, and content types in real-time. This dynamic approach enables the system to adjust ad selections based on varying page characteristics and user interactions, allowing consistent content-matching performance across diverse audiences and platforms.
3Productivity
If advertisements are selected in real-time based on user activities and demographics, then user response rate is improved, but system complexity increases
Solution Approach 1:
The patent performs preliminary analysis and categorization of web page content, metadata, and contextual information in advance of ad selection. By pre-processing and structuring content data, the system reduces the computational complexity required for real-time ad selection while maintaining the ability to respond quickly to user activities and deliver relevant ads.
4Adaptability or versatility
If demographic information is collected through cookies and tracking, then ad personalization is improved, but user privacy concerns increase
Solution Approach 1:
The patent uses content-based signals from web pages, metadata, and contextual information as an intermediary that enables ad personalization without directly accessing or storing sensitive user demographic data. This intermediary approach maintains privacy by substituting content-based targeting for user-data-based targeting while achieving similar personalization effects.
Data Source
AI summary
Systems and methods are provided for selecting advertisements for display over a communications network and, more particularly, to systems and methods for selecting video advertisements for display within internet web pages based on relevance, bid price, past performance, or a combination thereof.