Relationship Inference for Spam Filtering and Content Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing communication systems lack effective methods to dynamically manage and control interactions based on user relationships, leading to issues with spam filtering and inappropriate content exposure, especially for teen or child accounts.
Innovation Solution
A system that maintains a list of known individuals by inferring relationships through user actions and indicia, using this list to filter communications, restrict access, and manage spam, with features like white lists, black lists, and parental controls to ensure appropriate content delivery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If spam filtering is implemented to block unwanted communications, then harmful factors are reduced, but legitimate communications from unknown senders may be incorrectly blocked
Solution Approach 1:
The patent introduces an intermediary classification system that categorizes senders into different groups (known, unknown, spam, legitimate) based on user actions and communication patterns. This intermediary classification layer allows the system to make informed decisions about message routing, reducing false positives while maintaining spam filtering effectiveness.
Solution Approach 2:
The system implements feedback mechanisms where user actions (reading, deleting, reporting spam) are continuously monitored and used to update sender classifications. This feedback loop enables the system to learn from user behavior and improve its filtering accuracy over time, distinguishing legitimate communications from spam more effectively.
2Object-affected harmful factors
If parental controls are implemented to filter inappropriate content for teen or child accounts, then harmful factors are reduced, but communication freedom is restricted
Solution Approach 1:
The patent applies different filtering levels and rules to different user accounts based on their characteristics (teen, child, adult). Parental controls are selectively applied only to appropriate account types, allowing communication freedom for adult accounts while providing targeted protection for younger users without unnecessarily restricting their communication freedom.
Solution Approach 2:
The filtering system is dynamic and adaptable, allowing parents to adjust control levels as children mature. The system can evolve from strict filtering to more permissive modes based on user age, behavior patterns, and parental preferences, balancing protection with communication freedom over time.
3Object-affected harmful factors
If a list of known individuals is maintained and used to control communications, then harmful factors are reduced, but device complexity increases
Solution Approach 1:
The system implements self-service automation where the list of known individuals is automatically updated based on user actions and communication patterns. Users don't need to manually manage the list; instead, the system learns and updates it autonomously, significantly reducing management complexity while maintaining security.
Solution Approach 2:
The system performs preliminary classification and organization of contacts before communications occur. By pre-establishing the list of known individuals and their trust levels, the system simplifies real-time decision-making about message routing and filtering, reducing the complexity of on-the-fly judgments.
4Adaptability or versatility
If user actions are monitored to infer relationships, then adaptability is improved, but loss of information increases due to privacy concerns
Solution Approach 1:
The system extracts only the minimum necessary information from user actions to infer relationships, rather than monitoring all user activities. By selectively extracting relevant signals (email interactions, communication patterns) while leaving other private information untouched, the system maintains relationship inference accuracy while preserving user privacy.
Data Source
AI summary
The people a user is presumed to know or be associated with may be determined using a number of techniques. For example, people a user knows may be inferred based on a combination of two or more user actions, each of which separately support an inference that the person is associated with the user. This information about people that the user knows is used in relation to the user's communications.