Unsupervised Cluster Prioritization via Dissimilarity Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Telecommunications companies face challenges in effectively identifying and marketing products to customers due to the complexity of customer characteristics, leading to overwhelming generic marketing approaches and sub-optimal cluster selection in customer data analysis.
Innovation Solution
A systematic approach is introduced to identify and order interesting and important groups within a dataset by defining a reference cluster, computing dissimilarity measures using methods like Kullback-Leibler divergence, and visually representing clusters and attributes to prioritize and select key characteristics, enabling objective cluster ordering and visualization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If mass marketing product offerings are provided to customers, then marketing coverage is improved, but product purchase likelihood and customer engagement deteriorate
Solution Approach 1:
The patent segments customers into distinct clusters based on their characteristics and behaviors. By dividing the customer base into homogeneous groups (clusters), the system enables targeted marketing approaches that improve product purchase likelihood while maintaining comprehensive coverage. Each cluster can receive customized marketing offerings tailored to their specific needs and preferences.
Solution Approach 2:
The patent applies local quality by customizing marketing strategies for different customer clusters. Instead of uniform mass marketing, the system identifies and emphasizes cluster-specific characteristics and preferences to create localized marketing approaches. This allows each cluster to receive relevant product offerings that match their individual needs, thereby improving engagement and purchase likelihood.
2Device complexity
If generic marketing offerings are broadcast to all customers, then marketing simplicity is improved, but customer engagement and relevance deteriorate
Solution Approach 1:
The system segments customers into meaningful clusters that reveal underlying patterns in customer behavior and characteristics. This segmentation enables the system to maintain operational simplicity while achieving high relevance by processing customer data into distinct groups that can be targeted with appropriate marketing strategies.
Solution Approach 2:
The patent employs self-service through automated cluster identification and prioritization. The system automatically analyzes customer data, identifies clusters, determines their importance, and ranks them without requiring manual intervention. This automation maintains marketing approach simplicity while significantly improving customer engagement through personalized targeting.
3Measurement precision
If customers are described by a large number of characteristics, then customer description accuracy is improved, but classification and identification difficulty increases
Solution Approach 1:
The patent extracts and identifies the most important characteristics from the large set of available customer attributes. By extracting only the essential features that define meaningful clusters, the system maintains high customer description accuracy while reducing classification complexity. The extraction process automatically selects and emphasizes the most discriminative characteristics for each cluster.
Solution Approach 2:
The patent introduces cluster identification as an intermediary step between raw customer characteristics and final classification. This intermediary process organizes and prioritizes characteristics based on their importance in defining meaningful customer segments. The intermediary cluster structure simplifies the classification task while preserving the accuracy needed for effective customer identification.
4Measurement precision
If manual cluster selection and prioritization is performed, then cluster importance assessment accuracy is improved, but time consumption and labor requirements increase
Solution Approach 1:
The system performs self-service by automatically identifying, assessing, and prioritizing clusters without manual intervention. The automated process evaluates cluster importance based on predefined criteria and ranks clusters accordingly, eliminating time-consuming manual analysis while maintaining high assessment accuracy through systematic algorithmic evaluation.
Solution Approach 2:
The patent replaces manual mechanical cluster evaluation with automated computational methods. The system uses algorithms to automatically assess cluster characteristics, determine importance, and generate rankings. This substitution of mechanical manual processes with automated computational systems significantly reduces time consumption while preserving or improving assessment accuracy through consistent, objective criteria.
Data Source
AI summary
Techniques are disclosed that automatically identify and order the most differentiated clusters from a given collection of clusters within a dataset. A measure of dissimilarity is computed for each cluster from a defined reference cluster, and the clusters are ordered according to the chosen dissimilarity. At least N clusters are selected as the most differentiated clusters relative to the defined reference. Within each cluster, the top-M most distinguishing cluster attributes can be automatically identified by an analogous process that computes the dissimilarity of each cluster attribute to its corresponding attribute in the reference cluster, and orders the attributes by dissimilarity. This then allows for automatic surfacing of what it is about a cluster that differentiates its members relative to the population as a whole, and to provide insight on what action or treatment might be made to address that specific segment of the underlying population.