Generative UI Workflows for Personalized Mobile Page Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing user interface (UI) workflows in digital applications are standardized and fail to accommodate individual user preferences, leading to unsatisfactory experiences and increased support burdens.
Innovation Solution
A system utilizing large language models (LLMs) to dynamically generate personalized UI workflows by analyzing user behavior and preferences, allowing for custom interface elements and navigation structures based on user context profiles, with iterative feedback mechanisms for refinement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If standardized UI workflows are used, then device complexity is reduced and ease of manufacture is improved, but adaptability to individual user preferences deteriorates
Solution Approach 1:
The system performs preliminary actions by pre-defining multiple UI workflow templates with different characteristics (simplified, standard, complex) before the user interacts with the application. The LLM analyzes user preferences in advance and selects or customizes the appropriate workflow template, avoiding the need to create customized workflows from scratch during user interaction.
Solution Approach 2:
The UI workflow is made dynamic by allowing it to change based on user preferences. The system uses LLM to analyze user behavior and preferences in real-time, then dynamically adjusts the workflow complexity and characteristics to match the user's needs, transitioning between predefined templates or generating custom workflows as required.
2Adaptability or versatility
If personalized UI workflows are generated for each user, then adaptability to user preferences is improved, but device complexity and design costs increase
Solution Approach 1:
The system achieves universality by creating a single multi-functional platform that can serve all users with different preferences. Instead of designing separate customized workflows for each user, the LLM-powered system can generate appropriate workflows for any user based on their preferences, making the system adaptable to all users through a single unified architecture.
Solution Approach 2:
The system uses copying by creating UI workflows based on predefined templates that can be replicated and adapted. The LLM generates customized workflows by copying and modifying template structures according to user preferences, rather than creating entirely new workflows from scratch, thereby reducing design complexity and costs.
3Productivity
If LLM-based generation is used, then productivity in generating personalized UI workflows is improved, but use of energy and computational resources increases
Solution Approach 1:
The system applies partial action by using the LLM only when necessary - specifically when user preferences indicate a need for personalized workflows or when predefined templates are insufficient. For standard user preferences, the system uses simpler template-based generation, reserving the computationally intensive LLM generation for cases where it provides genuine value.
Solution Approach 2:
The workflow generation process is segmented into multiple stages: first, the system analyzes user preferences and selects an appropriate predefined template; second, the LLM is invoked only for generating or customizing specific portions of the workflow based on user needs; third, the generated workflow is assembled and deployed. This segmentation reduces overall computational resource consumption.
Data Source
AI summary
Systems and methods are disclosed comprising instructions to identify a user of a mobile device based on an interaction of the user on a communications network, retrieve from a database an aggregated history of activity of the user with the service on the communications network, receive a request to access a first page on the mobile device belonging to a set of pages that is formed of page elements that are linkable to each other as a sequence of pages, and in response to the request to access the first page on the mobile device, dynamically select a second page of the set of pages, generate a personalized page element for the second page, integrate the personalized page element into the second page to generate a personalized second page, and link the personalized second page to the first page, the second page being accessible as a next page of the first page.