Audience Selection via Knowledge Graph Expansion
Solution Overview
Problem
Existing content distribution systems struggle to effectively reach a desired audience for advertisements and sponsored content, as the number of directly interested users may not meet the specified reach, leading to inefficiencies in audience selection and ad placement.
Innovation Solution
The system uses a combination of social graphs and knowledge graphs to identify directly interested users and expand the audience by selecting users who have expressed interest in related entities, based on relationships and implicit interests, allowing for the selection of additional audience members to meet the desired reach.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system only selects directly interested users for content distribution, then the precision of audience selection is improved, but the reach (quantity of audience) deteriorates
Solution Approach 1:
The patent segments the audience selection process into two distinct phases: first identifying directly interested users based on explicit expressions of interest, then separately identifying additional audience members based on implicit interest through social connections. This segmentation allows the system to maintain precision for the core audience while expanding reach through a systematic secondary selection process.
Solution Approach 2:
The patent introduces socially connected users as an intermediary layer between directly interested users and the final audience selection. These intermediaries serve as a bridge, allowing the system to expand reach while maintaining a degree of precision through the social connection filter. The intermediaries help transfer the interest signal from direct users to potential additional audience members.
2Quantity of substance
If the system expands audience selection to include users interested in related entities, then the reach is improved, but the precision of audience selection deteriorates
Solution Approach 1:
The patent implements a dynamic, multi-tiered audience selection process where the system can adjust the balance between precision and reach based on distribution criteria. The selection process is not static but adapts by considering multiple factors including social connection strength, entity relationships, and expressed interest levels, allowing flexible optimization of the precision-reach tradeoff.
Solution Approach 2:
The patent changes the selection parameters from a single binary interest indicator to a multi-dimensional assessment including direct interest expressions, social connection strength, entity relationship types, and implicit interest signals. By changing these parameters, the system can expand reach while maintaining precision through weighted evaluation of multiple factors rather than relying on a single criterion.
3Measurement precision
If the system uses only explicit interest expressions for audience selection, then the precision is improved, but the complexity of the selection process increases due to insufficient data
Solution Approach 1:
The patent makes the interest data system multi-functional by enabling it to serve both precision selection (through direct interest expressions) and reach expansion (through social connections and entity relationships). The same data infrastructure supports multiple selection strategies, reducing the need for separate complex systems while maintaining precision through the primary direct-interest filter.
4Productivity
If the system distributes content to meet desired reach, then the productivity is improved, but the loss of information about user interest specificity increases
Solution Approach 1:
The patent performs preliminary classification of audience members into different tiers (directly interested, socially connected, additional audience) before final content distribution. This preliminary action allows the system to maintain information about user interest specificity through structured categorization, enabling efficient distribution to meet reach goals while preserving knowledge about the nature and strength of each user's interest.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting an audience for content are disclosed. In one aspect, a method includes receiving, from a content item provider, a request to distribute content items to users that have been deemed interested in a particular entity. First users that have expressed an interest in the particular entity are identified. Expansion entities for the particular entity are identified in a knowledge graph. At least one of the expansion entities can be connected to the particular entity by a relationship path. Second users are identified. The second users are deemed interested in the particular entity based on the second users having expressed an interest in an expansion entity. The content items are provided to at least a portion of the first users and at least a portion of the second users.