Personalized Cross-App Action Recommendations From User Behavior
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile applications lack personalized and efficient content and action recommendations, often requiring manual user input and tedious processes like augmented reality visualization, and virtual agents are limited by reliance on voice commands without considering previous user information.
Innovation Solution
A system utilizing a behavior analyzer that records user actions across multiple applications, develops personalized models based on these actions, and recommends follow-up actions or interacts as a virtual agent to assist users, leveraging components like location, vision, and text analyzers to enhance user experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If applications use hardcoded action recommendations, then action recommendations are provided, but the recommendations are not customized and lead to inferior user experience
Solution Approach 1:
The system performs preliminary actions by continuously tracking and recording user actions across applications in the background before recommendations are needed. This pre-collection of behavioral data enables personalized recommendations without requiring complex real-time analysis when the user interacts with the application.
Solution Approach 2:
The patent introduces an intermediary component (behavioral tracking system with machine learning model) that sits between user actions and recommendation generation. This intermediary automatically captures, analyzes, and stores user behavioral patterns, translating raw user actions into personalized recommendation insights without requiring direct complex processing at the point of recommendation.
2Productivity
If users manually enter instructions in applications, then precise user intent is captured, but the process is time-consuming and reduces productivity
Solution Approach 1:
The system implements feedback by continuously monitoring user actions across applications and using this information to refine and update the personalized model. The system learns from user responses to recommendations and adjustments to recommendations, creating a closed-loop feedback mechanism that improves accuracy over time without requiring explicit user input.
Solution Approach 2:
The system performs self-service by automatically tracking, analyzing, and generating recommendations based on user behavioral patterns without requiring manual user input. The machine learning model autonomously processes user actions and generates personalized recommendations, freeing users from the need to manually enter instructions while maintaining high accuracy in understanding user intent.
3Adaptability or versatility
If virtual agents rely only on voice commands, then simple interactions are enabled, but the agent cannot use previous user information to generate content recommendations
Solution Approach 1:
The patent implements universality by creating a centralized behavioral tracking and analysis system that serves multiple functions: it tracks user actions across applications, builds personalized models, generates content recommendations, and enables virtual agents to access historical user information. This multi-functional system eliminates the need for separate mechanisms for each function, reducing overall system complexity while enhancing adaptability.
Data Source
AI summary
A system for recommending actions on a device includes at least one processor configured to record actions, which are performed by a user on the device, across a plurality of applications present on the device. The processor develops a personalized model, which is specific to a user of the device, for recommending actions, wherein the personalized model is at least based on the recorded actions. The processor recommends a follow on action to be carried out on a second application after a first action is carried out on a first application based on the personalized model.