Ad Campaign Selection via Anonymized Log Similarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In computer networked environments, existing technologies face challenges in enhancing the performance of online advertisement campaigns while maintaining data protection and security, as direct sharing of Internet activity log data between content provider and publisher servers is undesirable due to privacy concerns.
Innovation Solution
A centralized data processing system maintains anonymized Internet activity log data, performing semantic and subject matter similarity analyses to select relevant content items for online ad campaigns, using bid value factors specified by content providers to bias content item selection without sharing actual log data between servers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If direct sharing of Internet activity log data between content provider and publisher servers is implemented, then ad campaign performance can be enhanced through better content selection, but data protection and security are compromised
Solution Approach 1:
A centralized data processing system acts as an intermediary between content providers and publishers. The system receives anonymized log data from content providers, performs semantic and subject matter similarity analyses, and returns content item selections to publishers without either party directly accessing the other's data. This mediator approach enables collaborative ad campaign optimization while maintaining data protection and security boundaries.
2Measurement precision
If anonymized log data is centralized for analysis, then content item relevance can be improved through semantic similarity analysis, but system complexity increases
Solution Approach 1:
The centralized data processing system performs multiple functions within a single platform: receiving and storing anonymized log data from multiple content providers, performing semantic similarity analysis between action types and relevance keywords, conducting subject matter similarity analysis, generating bid value factors, and selecting content items for various publishers. This multi-functional approach consolidates complexity into a single system rather than distributing it across multiple separate systems.
Solution Approach 2:
The system transforms raw anonymized log data into meaningful insights by applying parameter changes through semantic and subject matter similarity analyses. Bid value factors are generated based on these similarity measurements, converting qualitative data characteristics into quantitative parameters that drive content item selection decisions.
Data Source
AI summary
Systems and methods of selecting content items for an online ad campaign in a computerized network environment are described herein. The system can receive, from a first content provider, an event call. The event call can include an identifier and an action type. The system can receive, from a client device, a request for content, the client device associated with the identifier. The system can determine a similarity between the action type and a relevance keyword. The relevance keyword can be by a second content provider. The system can generate a bid value factor for a content item of the second content provider and a bid value factor based on the semantic similarity. The system can determine a bid value for the content item of the second content provider based on the bid value factor. The system can select, for transmission to the client computing device, the content item.