Shopping Intent Classification for Adaptive GUI Personalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing graphical user interfaces (GUIs) struggle to adapt to diverse user shopping intents and behaviors, leading to increased friction and inefficiency as they often rely on historical data and manual design expertise, failing to account for individual user needs and preferences.
Innovation Solution
A method and system that utilizes an intercept survey to capture user intent, collects usage data, segments sessions, trains a classification model, and modifies the GUI to align with predicted user intents, using machine learning to dynamically update the interface based on self-identified preferences and behaviors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single GUI is used to serve multiple users with diverse shopping intents, then the system maintains simplicity in deployment and management, but the user experience deteriorates due to increased friction and lack of personalization
Solution Approach 1:
The GUI dynamically adapts its layout, content, and functionality based on real-time classification of user shopping intent. The system transitions from a static single-interface approach to a dynamic multi-persona interface that automatically reconfigures based on detected user needs, resolving the contradiction between serving diverse users and maintaining system simplicity.
Solution Approach 2:
The system segments users into different intent categories (e.g., browsing, shopping, price comparison) and provides customized GUI experiences for each segment. This segmentation allows the single GUI to effectively serve multiple user types simultaneously without requiring separate systems for each user category.
2Measurement precision
If historical behavior data is used to infer user intent, then the system can operate without direct user input, but the measurement precision of user intent deteriorates due to inability to capture current shopping mindset
Solution Approach 1:
The system presents an intercept survey at the beginning of user sessions to capture current shopping intent before users engage with the GUI. This preliminary action ensures that the intent classification is based on users' current mindset rather than historical patterns, significantly improving measurement precision while requiring minimal time investment from users.
3Ease of operation
If manual design expertise is required to optimize GUI for different user types, then the system can provide personalized experiences, but the ease of manufacture deteriorates due to increased design and implementation complexity
Solution Approach 1:
The system automatically classifies user intent and selects appropriate GUI configurations without requiring manual design intervention for each user type. The classification model autonomously personalizes the interface based on detected shopping intent, eliminating the need for designers to manually create and manage multiple personalized interface versions while still providing customized experiences.
Data Source
AI summary
A method and system for updating a graphical user interface (GUI) based on self-identified user intents. The method performs the steps of: receiving a self-identified user intent from a user of a computing device with a user interface for a usage session; collecting usage data of the usage session of the user of the computing device and the user interface through an analytics service; labelling each usage session of the user with the self-identified user intent; training a classifier on the usage data of the user session and the self-identified user intent, the classifier for predicting an unclassified user intent of an unclassified user based on an unclassified user usage data and a self-identified unclassified user intent; assigning an assigned user intent to the unclassified user intent by using the classifier on the unclassified user usage data, and the self-identified unclassified user intent, the unclassified user now a classified user; modifying the user interface in response to the assigned user intent for facilitating the assigned user intent of the classified user; and selecting the user intent from a predetermined list of user intents.