Notification Balancing Using Similarity and Popularity Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computing systems face inefficiencies in managing device notifications, leading to resource wastage when users receive significantly different quantities, affecting user experience and resource utilization.
Innovation Solution
A system that processes interaction information using computational algorithms and Monte Carlo simulations to rank candidate entities based on similarity and popularity, adjusting notification quantities to optimize resource use.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If notifications are provided to all users based on device actions, then user engagement is improved, but computing and network resources are wasted when users receive significantly different quantities of notifications
Solution Approach 1:
The system changes the parameter of notification quantity dynamically based on user-specific similarity measures and popularity scores. Instead of uniform notification delivery, the system adjusts the number of notifications per user by calculating similarity between user actions and modifying notification quantities accordingly, thereby optimizing resource utilization while maintaining engagement.
Solution Approach 2:
The patent applies local quality by tailoring notification quantities to individual users based on their specific action patterns and similarity to other users. Each user receives a customized notification quantity determined by their local characteristics (action patterns, popularity), rather than a uniform approach, which reduces waste for users who would receive excessive notifications.
2Loss of energy
If notification quantities are balanced based on user similarity and popularity, then resource utilization is optimized, but system complexity increases due to computational algorithms and data processing
Solution Approach 1:
The system performs preliminary actions by pre-calculating similarity measures and popularity scores for users based on their action patterns stored in data structures. These pre-computed metrics are then used to determine notification quantities, avoiding the need for complex real-time calculations during notification delivery and reducing instantaneous system complexity.
Solution Approach 2:
The patent introduces intermediary data structures and computational layers that mediate between raw user action data and final notification delivery decisions. These intermediaries (similarity measures, popularity scores) simplify the complex relationship between user behaviors and notification quantities, making the system more manageable despite the underlying complexity.
3Productivity
If user actions are analyzed to determine similarity and popularity, then notification balancing is achieved, but data processing time and storage requirements increase
Solution Approach 1:
The system performs preliminary data processing by storing and pre-analyzing user action patterns in data structures before notification delivery is needed. This allows the system to quickly retrieve and utilize pre-computed similarity and popularity information rather than performing complex analyses in real-time, significantly reducing data processing time during notification delivery.
Solution Approach 2:
The patent extracts key features from raw user action data (such as action patterns, frequencies, and temporal characteristics) and stores them in optimized data structures. By extracting and pre-processing only the essential information rather than handling all raw data, the system reduces storage requirements and processing time while maintaining the accuracy of similarity and popularity calculations.
Data Source
AI summary
A device may obtain, from a data structure, interaction information regarding first actions performed by a first device of a first entity and regarding second actions performed by a second device of a second entity. The first actions and the second actions may cause notifications to be provided to devices of a plurality of users. The device may determine, based on the interaction information, a first measure of similarity between the first entity and a first candidate entity, a second measure of similarity between the first entity and a second candidate entity, a first measure of popularity of the first candidate entity, and a second measure of popularity of the second candidate entity. The device may provide candidate information regarding the first candidate entity and regarding the second candidate entity, based on the measures of similarity and the measures of popularity.