Affinity Scoring for Commerce User Interface Recommendations

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing commerce user interfaces and information systems rely on complex rule sets and customer segmentation, leading to inaccurate, incomplete, or irrelevant product and service recommendations due to their inability to quantify customer interest in specific products or categories, resulting in inefficient use of computer and network resources.

Innovation Solution

The implementation of affinity scoring techniques that process streaming interaction events in real-time to generate scores based on current, historical, and predicted user activities, allowing for personalized and relevant content recommendations by combining 'history' and 'future' scores with forward time decay, thereby enhancing user interface accuracy and efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If complex rule sets and customer segmentation are used to generate product recommendations, then the system can provide structured suggestions, but the accuracy and relevance of recommendations deteriorate due to inability to quantify customer interest

Engineering Contradiction:

Improverecommendation capabilityVSAvoidcustomer interest quantification

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent transforms qualitative customer behavior data into quantitative affinity scores through parameter transformation. Streaming events are converted into numerical values that represent customer interest levels, enabling precise measurement and comparison of customer preferences across different products and time periods.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces traditional rule-based mechanical recommendation systems with a data-driven affinity scoring mechanism. Instead of relying on pre-defined rules and customer segments, the system uses real-time event streaming and computational scoring to dynamically determine product recommendations.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If traditional recommendation systems are used, then processing is simpler, but information retrieval accuracy deteriorates leading to inefficient use of computer and network resources

Engineering Contradiction:

Improveinformation retrieval accuracyVSAvoidcomputer and network resource consumption

Core Design Contradiction:

ProductivityVSLoss of energy

Solution Approach 1:

The patent performs preliminary affinity score calculation and customer interest quantification before the actual information retrieval process. By pre-processing and scoring customer preferences in advance, the system reduces the computational burden during live information retrieval, improving efficiency while maintaining high accuracy.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses customer's own interaction history and behavior patterns to generate personalized recommendations without requiring external complex rule sets. The affinity scoring mechanism leverages self-generated data from customer interactions to autonomously determine relevant information, reducing resource consumption.

Inventive Principle:

Principle #25Self-service

3Ease of operation

If customer segmentation is applied to categorize customers, then the system can organize data structurally, but recommendation relevance deteriorates by providing imprecise untargeted information

Engineering Contradiction:

Improvedata organizationVSAvoidrecommendation relevance

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent transitions from uniform customer segmentation to localized, individualized affinity scoring. Instead of categorizing all customers into broad segments, the system calculates specific affinity scores for each customer based on their unique interaction patterns, providing locally optimized recommendations tailored to individual preferences.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent replaces static customer segmentation with dynamic affinity scoring that continuously updates based on real-time streaming events. Customer preferences are not fixed in predefined segments but dynamically recalculated as new interaction data arrives, ensuring recommendations remain relevant and precise.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS11205218B2Client user interface activity affinity scoring and tracking

Publication Date: 2021.12.21 BBY SOLUTIONS INC

AI summary

Techniques and system configurations for generating, updating, and customizing user interface content and functionality based on user activity and a customer level of interest (affinity) for a particular commerce information item are disclosed. In an example, electronic operations used for generating and updating output of a user interface includes: processing streaming events that represent user activity involving a commerce information item performed by a user in a user interface; identifying a current score of the user activity from the streaming events; generating an affinity score corresponding to the commerce information item, based on forward time decay of a history score and a future score; and providing output in a user interface based on the generated affinity score. With these techniques, engagement and affinity of particular products/services, product groups, and offers, can be measured and tracked to enable improved outputs in an electronic commerce website or app.