Purchase Category Association Engine for Multi-Category Recommendations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating purchase recommendations in online shopping struggle to effectively associate and recommend commercial objects across multiple categories, failing to capitalize on consumers' purchase patterns and preferences.
Innovation Solution
A computer-executable method that analyzes prior purchase data to generate category association scores, allowing for the recommendation of commercial objects in categories with high probability of interest based on consumers' purchase history and preferences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional recommendation methods are used, then implementation is simple, but recommendation accuracy across multiple categories is poor
Solution Approach 1:
The patent segments the recommendation system into distinct modules: a category association generation engine that analyzes purchase data to create category relationships, and a recommendation engine that uses these associations to generate recommendations. This segmentation allows complex multi-category analysis to be broken down into manageable components, improving recommendation accuracy while maintaining implementability.
Solution Approach 2:
The system performs preliminary action by pre-generating category association scores from historical purchase data before actual recommendations are needed. These pre-computed associations are stored and reused, allowing the recommendation engine to quickly leverage established category relationships without performing complex analysis in real-time, thus improving accuracy without proportionally increasing complexity.
2Adaptability or versatility
If purchase recommendations are generated without category associations, then processing is fast, but personalization and relevance are insufficient
Solution Approach 1:
The system performs preliminary analysis of purchase data to generate category association scores in advance, storing these associations for rapid retrieval during recommendation generation. This pre-computation enables personalized recommendations that adapt to individual consumer patterns without requiring extensive real-time processing, thus improving personalization capability while minimizing time loss.
Solution Approach 2:
The system uses historical purchase data as feedback to continuously refine category associations. By analyzing past consumer behavior patterns and updating category relationships based on this feedback, the system improves its personalization capability over time while the pre-computed nature of the associations keeps processing time manageable.
3Measurement precision
If category association analysis is performed for all consumers, then recommendation quality improves, but computational resources are excessive
Solution Approach 1:
The system segments the consumer base and applies category association analysis selectively rather than uniformly to all consumers. By identifying and focusing computational resources on consumers who would benefit most from personalized multi-category recommendations, the system improves recommendation precision for target users while reducing overall computational resource usage.
Solution Approach 2:
The system performs partial analysis by generating category associations only for relevant consumer segments or only for specific category pairs that demonstrate meaningful relationships in the data. This partial action approach maintains high recommendation precision for targeted users while avoiding the excessive computational resource consumption that would result from analyzing all possible consumer-category combinations.
Data Source
AI summary
Embodiments provide a computer-executable method, computer system and non-transitory computer-readable medium for programmatically generating an association among two or more purchase categories based on purchase data of a plurality of consumers. The method includes programmatically accessing, from a dataset via a network device, prior purchase data associated with purchases of a plurality of commercial objects by a plurality of consumers. The method also includes programmatically identifying a plurality of categories associated with the plurality of commercial objects. The method also includes, for each consumer in the plurality of consumers, programmatically generating a total number of purchases by the consumer in each category in the plurality of categories. The method further includes generating, using a processor of a computing device, a category association score between each pair of categories in the plurality of categories by programmatically analyzing similarities among the total numbers of purchases in the plurality of categories for the plurality of consumers.