Heterogeneous Data Network Merging for Cross-Retailer Recommendations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing recommender systems are limited in providing recommendations for alternate products or services beyond those offered by the same retailer, failing to leverage diverse data sources effectively.
Innovation Solution
A method and system that combines matrices of merchant location interactions and product data to generate a third matrix, enabling recommendations of products and merchants across different locations, using payment card transaction data while maintaining user anonymity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If recommender systems use historical data from a single retailer to determine user preferences, then recommendations can be generated within that retailer's ecosystem, but the system cannot provide recommendations for alternate products or services from different retailers
Solution Approach 1:
The patent merges heterogeneous data from multiple independent networks (payment card network, social network, mobile device network) into a unified recommendation system. Each network contributes its own data corpus, and these are combined through matrix operations to create comprehensive user preference profiles that enable cross-retailer recommendations while preserving the anonymity and structural integrity of each source network.
2Adaptability or versatility
If multiple heterogeneous data networks are merged to provide diverse recommendations, then recommendation versatility improves, but the complexity of data integration and processing increases
Solution Approach 1:
The patent segments the complex data merging process into distinct matrix operations: generating interaction matrices from each heterogeneous data source separately, then combining these matrices through standardized operations. This segmentation allows each network's data to be processed independently according to its own structure, reducing the complexity of integration while maintaining the ability to provide comprehensive cross-retailer recommendations.
Data Source
AI summary
A method and system for merging heterogeneous data types are provided. The method includes receiving a first corpus of first data, the first data includes an indicator of an interaction between a first element of the first corpus of first data and a second element of the first corpus of first data, receiving a second corpus of second data, the second data includes an indication of an interaction between a third element of the second corpus of second data and a fourth element of the second corpus of data, and generating a third matrix using correlations of the first and second elements with correlations of the third and fourth elements.