<![CDATA[Market segmentation represents a critical challenge in local product marketing, particularly when addressing complex consumer behavior patterns and uncertain classification environments in today's digital economy. This research develops and validates a hybrid model integrating K-Means Clustering with Dempster-Shafer theory to enhance segmentation accuracy and reliability for local product markets. The K-Means algorithm groups consumers based on demographic, psychographic, and behavioral characteristics, while Dempster-Shafer theory quantifies uncertainty and provides confidence measures for segment assignments. Data collection involved comprehensive consumer surveys and transaction records from 2,847 participants across multiple local product categories over a 12-month period. The hybrid model achieved superior performance with 87.5% accuracy, 85.3% precision, 86.1% recall, and 85.7% F1-score, representing improvements of 5.4% over standard K-Means and 8.2% over hierarchical clustering methods. Four distinct market segments were identified: Young Urban Professionals (28%), Value-Conscious Families (35%), Traditional Loyalists (22%), and Digital Natives (15%), each exhibiting unique purchasing patterns, digital engagement levels, and price sensitivity characteristics. Cross-validation yielded a consistency score of 0.91 with segment stability demonstrated through 8.3% churn rate and conflict measure K = 0.12, indicating substantial agreement among evidence sources. The methodology successfully addresses uncertainty in consumer classification while providing actionable insights for targeted marketing strategies, pricing optimization, and customer retention programs. Local product marketers can implement this framework to develop evidence-based marketing approaches that accommodate both traditional and digital consumer preferences, enabling competitive positioning in increasingly complex market environments. The research establishes a scalable and practical solution for small to medium enterprises seeking sophisticated market analysis capabilities without requiring extensive computational infrastructure or technical expertise.]]>
Copyrights © 2025
