Recently Viewed Item Ranking Using ML Engagement Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems display recently viewed items in reverse-chronological order, assuming user interest, which is not always accurate, leading to suboptimal engagement.
Innovation Solution
Implement a system using machine learning models to personalize the ranking of recently viewed items based on user and item features, determining engagement scores and diversifying item categories to enhance user engagement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If recently viewed items are displayed in reverse-chronological order, then the system is simple to implement, but user engagement is suboptimal because it assumes user interest without actual personalization
Solution Approach 1:
The system changes the ranking parameters from simple reverse-chronological order to a personalized ranking based on multiple factors including time decay, user behavior patterns, item attributes, and contextual information. This allows the system to maintain simplicity in implementation while significantly improving user engagement accuracy through data-driven parameter adjustments.
Solution Approach 2:
The patent replaces the mechanical/simple chronological ordering system with a machine learning-based personalization system. The ML model processes user behavior data, item attributes, and contextual information to generate personalized rankings, substituting the basic mechanical sorting mechanism with an intelligent system that adapts to individual user preferences.
2Reliability
If a machine learning model is used to personalize item ranking, then user engagement is improved, but system complexity increases
Solution Approach 1:
The system segments the personalization process into distinct components: data collection modules, feature engineering components, machine learning modeling layers, and ranking optimization layers. This segmentation allows the complex ML system to be managed through modular components, each handling specific aspects of personalization, thereby reducing overall system complexity while maintaining high engagement accuracy.
Solution Approach 2:
The system performs preliminary actions by pre-processing user behavior data, item attributes, and contextual information before feeding them into the machine learning model. Features are engineered and prepared in advance, which simplifies the actual modeling process and reduces the computational complexity during real-time ranking operations.
3Ease of manufacture
If recently viewed items are shown in reverse-chronological order, then the display is straightforward, but item diversity is limited and conversion likelihood is reduced
Solution Approach 1:
The system applies local quality by tailoring the ranking and display of items to individual user characteristics, preferences, and behaviors. Instead of a uniform reverse-chronological display, each user receives a customized ranking that prioritizes items most likely to convert based on their specific patterns, thereby increasing conversion likelihood while maintaining ease of display implementation through automated personalization.
Data Source
AI summary
A method can include determining one or more features associated with a user and also associated with recently viewed items for the user. The method further can include determining, at least in part by a machine learning model, a respective engagement score for each of the recently viewed items based on one or more first features of the one or more features. The one or more first features can be determined by a correlation analysis of the one or more features in a training process of the machine learning model. The method additionally can include ranking the recently viewed items based on the respective engagement score for each of the recently viewed items. The method also can include transmitting, via a computer network to a user device of the user, the recently viewed items, as ranked, for display on the user device. Other embodiments are disclosed.