Targeted Customer Clustering for Mixed Attribute Segmentation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Customer segmentation methods face challenges in handling mixed attribute types, particularly numerical and categorical attributes, as existing clustering techniques require manual cluster labeling, which is tedious and inefficient for business users.

Innovation Solution

A system and method that convert both categorical and numerical attributes into a same-scale numerical format using a target attribute, enabling any clustering algorithm to identify clusters without manual labeling, allowing for efficient segmentation of customers based on sales data and controlling robotic inventory mechanisms.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual cluster labeling is used for customer segmentation, then classification accuracy is improved, but operation time and complexity increase significantly

Engineering Contradiction:

Improveclassification accuracyVSAvoidoperation time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs automatic cluster labeling through unsupervised learning algorithms that self-organize customer data into segments based on similarity metrics, eliminating the need for manual intervention. The algorithm automatically identifies patterns and assigns labels to clusters, making the system self-sufficient in the labeling process.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces the mechanical manual labeling process with an automated computational system using clustering algorithms. The system substitutes human operators with machine learning models that process customer data and generate cluster labels automatically, significantly reducing operation time while maintaining segmentation quality.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If manual cluster labeling is used for customer segmentation, then segment quality is improved, but ease of operation deteriorates

Engineering Contradiction:

Improvesegment qualityVSAvoidease of operation

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The clustering system performs automatic segment identification and labeling without requiring business users to manually define clusters. The algorithm self-organizes the data and produces segmented results that are ready for use, making the process accessible to users without specialized knowledge of cluster labeling techniques.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent introduces an automated clustering algorithm as an intermediary between raw customer data and segmented results. This intermediary system handles the complex task of cluster identification and labeling, shielding business users from technical complexities while delivering high-quality segmented outputs.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Productivity

If clustering algorithms are used without manual labeling, then productivity is improved, but measurement precision may deteriorate

Engineering Contradiction:

Improvesegmentation efficiencyVSAvoidsegment quality

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system replaces manual cluster labeling with automated unsupervised learning algorithms that process customer data efficiently. The algorithms use similarity metrics and distance calculations to automatically identify and label clusters, maintaining segmentation quality while dramatically improving processing speed and productivity.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent transforms the segmentation approach by changing from manual parameter-based labeling to algorithm-driven automatic clustering. The system uses mathematical parameters such as distance metrics and similarity thresholds to automatically determine cluster boundaries and labels, achieving both high productivity and maintained segment quality through optimized algorithmic parameters.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11803868B2System and method for segmenting customers with mixed attribute types using a targeted clustering approach

Publication Date: 2023.10.31 ORACLE INT CORP

AI summary

Systems, methods, and other embodiments are disclosed that are configured to segment customers using mixed attribute types. In one embodiment, a computerized data structure is read. The computerized data structure has numerical demographic attribute data, categorical demographic attribute data, and target attribute data that is associated with customers and is stored in a computerized memory. The numerical demographic attribute data and the categorical demographic attribute data are converted to a same numerical scale, based at least in part on the target attribute data, to form congruent attribute data in a format that is compatible with performing a cluster analysis on the congruent attribute data. The cluster analysis is performed on the congruent attribute data to generate segmented customer data representing a segmentation of the customers. The segmented customer data may be used to control at least one enterprise function performed by a computerized management system.