Social Network Advertising Label Propagation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Social networks face challenges in accurately targeting online advertisements to users with incomplete or sparse user profiles, as well as mapping colloquialisms and non-advertiser keywords to relevant advertising keywords.
Innovation Solution
A computer-implemented method and system that generates content-based keywords for users in social networks by labeling nodes with advertising labels based on content generated by users and neighboring nodes, allowing for the inference of user interests and targeting of advertisements even when users do not provide sufficient information in their profiles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If user profiles are kept simple and sparse, then ease of user operation is improved, but advertising targeting accuracy deteriorates
Solution Approach 1:
The patent introduces an intermediary inference system that uses social network relationships as mediators to transfer interest information from users with complete profiles to users with sparse profiles. The system acts as a bridge, propagating advertising labels through the social network graph to enable accurate targeting without requiring detailed user input.
Solution Approach 2:
The system implements feedback loops where advertising labels are propagated through social relationships and then used to improve targeting accuracy. The iterative process allows the system to continuously refine user interest profiles based on social network patterns and advertising interactions.
2Adaptability or versatility
If content-based keywords are generated for all users, then advertising targeting coverage is improved, but system complexity increases
Solution Approach 1:
The patent segments the user population into different groups based on profile completeness and social network characteristics. Users are processed differently depending on their profile status, with the system focusing computational resources on users who need inference while leveraging existing data for users with complete profiles.
Solution Approach 2:
The system applies partial action by generating advertising labels only for users who need them (those with sparse profiles), rather than processing all users uniformly. This selective approach reduces overall system complexity while maintaining comprehensive coverage.
3Measurement precision
If social relationships are used to infer user interests, then advertising targeting accuracy is improved, but data processing requirements increase
Solution Approach 1:
The system performs preliminary actions by pre-computing and storing advertising labels for users with complete profiles before the inference process begins. This pre-processing reduces the computational burden during the actual inference phase, as the system only needs to propagate existing labels rather than compute everything from scratch.
Solution Approach 2:
The patent uses copying by replicating advertising labels from users with complete profiles to their social connections. Instead of re-computing interest profiles, the system copies proven labels through the social network graph, significantly reducing data processing requirements while maintaining accuracy.
Data Source
AI summary
The subject matter of this specification can be embodied in, among other things, a method that includes generating content-based keywords based on content generated by users of a social network. The method includes labeling nodes comprising user nodes, which are representations of the users, with advertising labels comprising content-based keywords that coincide with advertiser-selected keywords that are based on one or more terms specified by an advertiser. The method also includes outputting, for each node, weights for the advertising labels based on weights of advertising labels associated with neighboring nodes, which are related to the node by a relationship.