Email Filtering via User Interaction Metadata Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing email filtering systems face challenges in distinguishing between unsolicited bulk email and gray mail, as global filtering methods often fail to account for user preferences, leading to inaccurate classification and user dissatisfaction.
Innovation Solution
A method that collects metadata on user interactions with email messages, allowing for personalized filtering at both the user and global levels, where personalization data is used to modify the global filter's decisions on message delivery, disposal, or routing, and is fed back to train the filter for future interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If global filtering is applied to all users independently of user preference, then the system can maintain a unified filtering approach, but the accuracy of filtering for individual users deteriorates due to inability to account for personal preferences
Solution Approach 1:
The patent segments the filtering system into two distinct components: a global filter that applies unified rules to all users, and a personalization filter that applies user-specific preferences. This segmentation allows each component to specialize - the global filter handles general spam patterns while the personalization filter handles individual user preferences, thereby resolving the contradiction between ease of implementation and filtering accuracy.
Solution Approach 2:
The patent merges the global filter and personalization filter into a unified filtering pipeline where both filters work together on the same email stream. The global filter processes emails first, then the personalization filter applies user-specific rules to the results. This combination allows the system to maintain both the simplicity of global filtering and the accuracy of personalized filtering simultaneously.
2Loss of information
If filter training relies on explicit user classifications via user interface, then the system can obtain user feedback, but the responsiveness of the mail system deteriorates due to inherent delay in receiving feedback
Solution Approach 1:
The patent implements preliminary action by proactively analyzing user interactions with emails (such as reading patterns, forwarding, deleting) and automatically converting these passive behaviors into training data before the user needs to manually classify anything. This preliminary collection and processing of feedback data eliminates delays and maintains system responsiveness while still gathering comprehensive user preferences.
Solution Approach 2:
The system performs self-service by automatically interpreting user interactions with emails as implicit feedback without requiring active user participation through the interface. The system monitors and analyzes user behavior patterns autonomously, converting passive user actions into actionable training data, thereby eliminating the need for users to manually provide feedback while maintaining high system responsiveness.
3Adaptability or versatility
If feedback from user interactions is not incorporated into global level filters, then individual user preferences are respected, but the overall accuracy of the global filter deteriorates due to lack of learning from user behavior
Solution Approach 1:
The patent implements a feedback mechanism where user interaction data collected at the personalization level is aggregated and used to continuously train and improve the global filter. The system creates a feedback loop where individual user preferences inform global filtering rules, allowing the global filter to learn from collective user behavior patterns while still maintaining the ability to accommodate individual preferences through the personalization layer.
Data Source
AI summary
A computer implemented method for filtering unwanted bulk email in an email system and providing a positive user experience is provided. The method enables protection of email users from unsolicited bulk email using user-provided data on user interactions at both a user storage level and a global level with an email system. Metadata on user interactions with messages is collected. Messages are received by the system and evaluated using a global filter which assigns a score resulting in a message action. The action may be message delivery, message non-delivery or message routing, based on a score assigned by the global filter. When the message is delivered to user storage, the message may be examined relative to the metadata, and may alter the message action to an action different than the message action resulting from the score. Metadata for a plurality of users is returned to the global filter for use in making filtering future messages and modifies the global filter.