Electronic Catalog Item Set Generation with Diversity Adjustments
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current electronic catalog systems fail to effectively recommend complementary and diverse items to users, often leading to increased navigation complexity and missed opportunities for purchasing substitute or complementary products.
Innovation Solution
Implement a system that analyzes user interaction data to form sets of items, including both similar and different types, based on co-purchase behaviors, and uses algorithms to generate recommendations that include pairwise scores, filtering, boosting, and diversity adjustments to present a curated list of items.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If current electronic catalog systems display only products of the same type matching the search query, then search results are precise and relevant, but users miss opportunities to purchase complementary or substitute products and navigation complexity increases
Solution Approach 1:
The system segments product recommendations into distinct categories: same-type products, complementary items, and substitute products. This segmentation allows the system to present diverse recommendation types separately, making it easier for users to understand and navigate the recommendations without overwhelming them with a mixed list of all possible products.
Solution Approach 2:
The system introduces an intermediary layer of curated item sets that mediate between the user's search query and the full product catalog. These curated sets act as a filtered intermediary, presenting only the most relevant complementary and substitute items alongside exact matches, thereby reducing navigation complexity while maintaining information completeness.
2Quantity of substance
If the system displays only exact matches for the search query, then result precision is high, but item diversity is low and users are not exposed to complementary products
Solution Approach 1:
The system applies local quality by differentiating the presentation of recommendation types. Exact matches are highlighted with higher prominence and precision, while complementary and substitute items are presented with appropriate contextual labeling. This allows each recommendation type to maintain its specific quality characteristics while contributing to overall diversity.
Solution Approach 2:
The system adds a new dimension to search results by introducing recommendation metadata that categorizes items as exact matches, complements, or substitutes. This dimensional enhancement allows users to perceive both the relevance (precision) and diversity (quantity of different types) simultaneously, resolving the trade-off between these two dimensions.
3Adaptability or versatility
If the system analyzes user interaction data to form item sets and generate recommendations, then item diversity and relevance improve, but system complexity increases
Solution Approach 1:
The system performs preliminary analysis of user interaction data to pre-compute item sets and recommendation scores before the user actually searches or views products. This preliminary action stores processed information in a structured format, reducing the computational complexity during actual user interactions while maintaining high adaptability and customization.
Data Source
AI summary
Systems and methods are provided for determining pairwise scores for items in an electronic catalog that is stored in a storage device communicatively coupled to a server, and determining when a pair of items in the electronic catalog is valid. A sum of the pairwise scores by activity type for a plurality of activity types may be determined, and the pairwise scores for the plurality of activity types may be summed. A sorted list of items that compliment an anchor item of the electronic catalog may be generated. Items from the generated sorted list may be filtered, and an item ranking in the filtered list may adjusted. Consistency between recommendations of the items in the adjusted ranking list may be determined, final recommendations of the items based on the determined consistency between recommendations may be selected and transmitted for display.