Behavioral Site Lists for Cookieless Content Targeting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing programmatic platforms face challenges in rapidly and accurately evaluating presentation opportunities for auxiliary content on wireless communication devices, especially when pseudonymous subscriber identities are absent, and rely heavily on cookies which are being deprecated, leading to inefficient targeting and resource waste.
Innovation Solution
A system generates behavioral site lists associated with interest categories by analyzing SDK data and broadband data to identify pseudonymous subscriber identities' interests, allowing programmatic platforms to target content without relying on cookies or pseudonymous identities, using machine learning models to adapt and update lists based on evolving subscriber behaviors and interests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If programmatic platforms rely on cookies and pseudonymous subscriber identities for evaluating presentation opportunities, then targeting accuracy is improved, but system reliability deteriorates as cookies are deprecated and pseudonymous identities become unavailable
Solution Approach 1:
The patent introduces behavioral site lists as an intermediary mechanism that bridges the gap between deprecated tracking methods (cookies, pseudonymous identities) and the need for accurate content targeting. These lists serve as a mediator that connects presentation opportunities with relevant content based on aggregated behavioral patterns rather than individual user tracking, maintaining targeting accuracy while improving system reliability in a post-cookie environment
Solution Approach 2:
The system creates simplified copies of user interest profiles through behavioral site lists that capture essential targeting information without relying on deprecated identification methods. Instead of copying individual user data through cookies, the system creates aggregated behavioral patterns that replicate the targeting functionality needed for content selection, enabling accurate content delivery without the unreliable cookie-based infrastructure
2Loss of time
If the system processes bid requests and generates behavioral site lists in real-time, then content targeting timeliness is improved, but computational complexity increases
Solution Approach 1:
The system performs preliminary actions by pre-processing behavioral data to create behavioral site lists before bid requests arrive. This includes aggregating content interaction data, identifying behavioral patterns, and organizing content into categorized lists in advance. When a bid request arrives, the system can quickly match it against pre-generated behavioral site lists rather than performing complex analysis in real-time, significantly reducing processing time while maintaining targeting accuracy
Solution Approach 2:
The patent segments the complex task of content targeting into distinct components: data collection, behavioral pattern identification, behavioral site list generation, and bid request matching. Each segment can be processed independently and optimized separately, reducing the computational burden on any single system component. The segmentation allows parallel processing of multiple bid requests against pre-generated behavioral site lists, improving throughput while managing complexity
3Measurement precision
If the system collects and analyzes extensive SDK and broadband data to generate interest category lists, then content relevance is improved, but data processing requirements increase
Solution Approach 1:
The system extracts only the essential behavioral signals from extensive SDK and broadband data that are most relevant for content targeting. Instead of processing all collected data, the system identifies and extracts key patterns such as content interaction frequency, time spent on content, and engagement metrics. This extraction approach maintains high content relevance by focusing on the most predictive behavioral indicators while significantly reducing the computational energy required to process the full dataset
Data Source
AI summary
A method of generating a plurality of behavioral site lists. The method comprises receiving a plurality of interest category lists by a behavioral site list generation application executing on a computer system, wherein each interest category list comprises a plurality of pseudonymous subscriber identities associated with the pre-defined interest category; receiving a plurality of data on subscriber visits to web sites and on application identities by the application; generating behavioral site lists by the application based on determining over-indexing of pseudonymous subscriber identities provided in the data on subscriber visits to web sites and on application identities in the interest category lists, wherein each behavioral site list comprises a plurality of web site addresses and a plurality of application identities that are deemed to have an above average association to subscribers with an interest in the interest category associated with the behavioral site list.