Messaging System Message Ranking via Playback Behavior Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing messaging systems fail to efficiently prioritize voicemail and other communications based on user interest, requiring significant user administration and lacking automation that accounts for user behavior with different callers.
Innovation Solution
Messages are ranked according to user actions with previous messages from the same sender, using header information and playback data to predict user interest, with higher rankings given to senders whose messages are more thoroughly listened to, and lower rankings to those barely accessed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If messages are prioritized based on user-defined groups and priorities, then message access efficiency is improved, but system complexity and user administration burden increase
Solution Approach 1:
The system automatically analyzes user behavior patterns with messages from different senders and autonomously determines message priority rankings without requiring user intervention. The system serves itself by learning from user actions (deleting, listening to, skipping messages) and automatically adjusting the presentation order of future messages, eliminating the need for manual group definitions and priority assignments.
Solution Approach 2:
The system continuously monitors and analyzes user feedback regarding message handling (deletion, playback completion, skipping) and uses this feedback to dynamically adjust message priority rankings. This closed-loop feedback mechanism enables the system to adapt to changing user preferences and improve message access efficiency over time without additional user administration.
2Ease of operation
If automated priority assignment is implemented, then user administration is reduced, but accuracy in predicting user interest decreases
Solution Approach 1:
The automated system achieves high prediction accuracy by continuously analyzing actual user behavior feedback (which messages are deleted, fully listened to, or skipped) and using this feedback to refine its priority ranking algorithm. The system learns from each user interaction to improve its predictions of user interest, maintaining accuracy without requiring user administration.
Solution Approach 2:
The system autonomously performs the complex task of analyzing user behavior patterns and adjusting message priorities without user involvement. It automatically tracks user actions across multiple messages from different senders, computes engagement metrics, and dynamically re ranks messages based on learned patterns, providing accurate predictions entirely through automated means.
3Device complexity
If messages are presented in received order, then system simplicity is maintained, but time spent on uninteresting messages increases
Solution Approach 1:
The system maintains simplicity by automatically determining message priority based on analyzed user behavior patterns without requiring complex user-defined criteria. It autonomously computes engagement metrics from user actions and directly applies these to message ranking, achieving time savings through automated intelligent sorting rather than user-configurable complex rules.
Solution Approach 2:
The system changes the ordering parameter from simple receipt time to a computed priority score based on user engagement behavior. By transforming the sorting criterion from a single parameter (receipt order) to a behavior-based priority parameter, the system reduces time on uninteresting messages while maintaining operational simplicity through automated parameter computation.
Data Source
AI summary
The present invention provides for the ranking of messages for presentation to a user based on user behavior with respect to previous messages received from ranked communication endpoints. In particular, the percentage of a message played back by a user can be used to assigned a rank for future messages received from the initiating communication endpoint associated with the subject message. Initiating communication endpoints associated with messages that the user has listened to all or a large proportion of are associated with a relatively high ranking that is applied to future messages from those communication endpoints. Initiating communication endpoints associated with messages that the user listens to only the message header or a small proportion of the message itself are assigned a relatively low ranking. Messages received as textual communications or by a unified messaging application may be ranked according to the order in which the user selects the messages for retrieval.