User Interest Determination via Weighted App Embedding Pooling

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing systems struggle to accurately determine user interests over time due to limited access to proprietary data from device manufacturers and restrictions on third-party access to user interaction data from various applications, hindering personalized content delivery.

Innovation Solution

A method and system that analyze application data from user devices to identify user embeddings, perform weighted mean pooling on selected app embeddings based on application update frequencies, and provide recommendations matching the determined user interests, bypassing proprietary information restrictions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If entities access proprietary data from device manufacturers and application sources to determine user interests, then the accuracy of user interest assessment is improved, but access restrictions and data availability limitations worsen

Engineering Contradiction:
Improveuser interest assessment accuracyVSAvoiddata accessibility
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent introduces an intermediary system that collects application data from multiple applications through standardized interfaces and aggregates it to determine user interests. This intermediary approach bypasses the need for direct access to proprietary data from individual applications or device manufacturers, resolving the contradiction between data accuracy and accessibility.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If third-party entities access user interaction data from applications, then personalized content delivery is improved, but application security and privacy restrictions worsen

Engineering Contradiction:
Improvepersonalized content deliveryVSAvoiddata access restrictions
Core Design Contradiction:
Adaptability or versatilityVSObject-affected harmful factors

Solution Approach 1:

The system enables applications to voluntarily contribute their own interaction data to the centralized collection system through self-service mechanisms. Applications opt-in to share data according to their own security policies, eliminating the need for third-party forced access while still enabling personalized content delivery through aggregated data.

Inventive Principle:
Principle #25Self-service

3Reliability

If device manufacturers retain exclusive access to collected data, then data security and control are improved, but user interest determination capability for other entities worsens

Engineering Contradiction:
Improvedata controlVSAvoiduser interest determination
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent creates a universal data collection system that serves multiple functions: it maintains data security through controlled access while enabling user interest determination for multiple entities simultaneously. The system acts as a multi-functional platform that benefits both data providers and content delivery entities without requiring exclusive access by any single party.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS20250014075A1System and Method for Determining User Interest Over Time Based on Application Data
Publication Date: 2025.01.09 GLANCE INMOBI PTE LIMITED
  • US20250014075A1 patent drawing
  • US20250014075A1 patent drawing
  • US20250014075A1 patent drawing

AI summary

Disclosed is a method and system for determining user interest of at least one user based on interactions of the at least one user with the user's user device and providing recommendations to the at least one user's user device based on the determined user interest of the at least one user. The method includes identifying user embeddings, by a first embeddings identifier module. The method includes identifying user interest representation, by the user interest identification module, based on identified user embeddings for determining user interest of the at least one user. Lastly the method includes providing recommendations to the at least one user's user device, by a recommendation module, based on determined user interest of the at least one user.