Target Expansion System for Messaging Relevance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing online social networking services face challenges in expanding message targets without sending irrelevant messages to members, as conventional methods often result in annoyance due to lack of relevance.
Innovation Solution
A target expansion system that determines outcome probabilities for messages and uses machine learning to identify and send messages to targets with a higher likelihood of positive responses, adjusting based on differential probabilities and weight factors to minimize negative responses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If messages are sent to all members, then message distribution coverage is improved, but member annoyance increases due to lack of relevance
Solution Approach 1:
The patent applies local quality by customizing message distribution to specific segments of the member population based on their characteristics, interests, and behaviors. Instead of uniform messaging to all members, the system tailors message delivery to relevant subsets, ensuring each member receives messages appropriate to their profile and preferences, thereby maintaining high coverage while avoiding annoyance from irrelevant content
Solution Approach 2:
The patent segments the member base into different groups based on various criteria such as demographics, interests, engagement patterns, and message responsiveness. This segmentation allows the system to identify and target specific subsets of members who are most likely to find each message relevant, thus expanding coverage to appropriate audiences while preventing annoyance from irrelevant messaging
2Object-affected harmful factors
If messages are sent to a limited target, then member annoyance is reduced, but message distribution coverage is restricted
Solution Approach 1:
The patent implements dynamic target expansion by continuously learning from member responses and interactions. The system starts with a conservative target set but dynamically expands the message distribution audience over time as it gains confidence in predicting member interest. This allows coverage to grow while maintaining relevance, as the expansion is driven by actual engagement data rather than static assumptions
Solution Approach 2:
The patent incorporates feedback mechanisms that monitor member responses to messages, including engagement metrics, click-through rates, and explicit feedback. This feedback loop allows the system to learn which members respond positively to which types of messages, enabling continuous optimization of the target audience. The feedback drives iterative expansion of the message distribution coverage while maintaining high relevance by incorporating real-world response data
3Measurement precision
If machine learning is used to determine message targets, then message relevance is improved, but system complexity increases
Solution Approach 1:
The patent introduces machine learning models as intermediary components that bridge the gap between raw member data and message distribution decisions. These models act as mediators that process complex member profiles, historical behaviors, and message characteristics to generate relevance predictions. By positioning the ML system as an intermediary layer, the patent manages complexity by encapsulating sophisticated algorithms within a standardized interface that integrates with existing message distribution infrastructure
Data Source
AI summary
This disclosure relates to systems and methods that include configuring a machine learning system to train on a plurality of messages transmitted to target groups of an online social networking service, determining a threshold differential and a weight value using responses to the plurality of messages, and send the input message to the target in response to a differential between the expected number of positive responses and the weight multiplied by the expected number of negative responses being greater than the threshold differential.