Variant-Aware Ranking Model for E-Commerce Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search systems in electronic catalogs only retrieve and present a single version of a product, failing to match variant items to search queries or user intents, leading to incomplete search results.
Innovation Solution
A system and method that utilize a variant-aware ranking model to independently rank variant items based on variant scores and relevancy scores, generating a set of recommended items for inclusion in an interface, allowing for better matching of user intent and search queries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If current search systems retrieve and present only a single version of a product, then the search results are simplified and easier to manage, but the completeness and relevance of search results deteriorate because variant items are not matched to search queries
Solution Approach 1:
The patent segments the single product representation into multiple variant items, each with distinct attributes (size, color, style). Instead of presenting one aggregated product, the system separates and presents individual variants that match different user intents, thereby maintaining simplicity while improving completeness.
Solution Approach 2:
The patent adds a new dimension to search results by introducing variant scoring and independent ranking. This creates an additional layer of organization beyond the single product level, allowing the system to present multiple variants ranked by their relevance to the search query, thus resolving the contradiction between simplicity and completeness.
2Ease of operation
If current search systems return only a single most-common variant of a product, then the presentation is consistent and straightforward, but the ability to match user intent deteriorates because different user needs are not addressed
Solution Approach 1:
The patent applies local quality by assigning different relevance scores to different variants based on their specific attributes and the search query. Each variant is evaluated independently with its own quality metric, allowing the system to present the most relevant variant for each user intent while maintaining straightforward presentation through consistent ranking criteria.
Solution Approach 2:
The patent introduces dynamic adaptability by making the ranking system responsive to different search queries and user intents. The variant-aware ranking model dynamically adjusts which variants are presented based on the specific search context, enabling the system to adapt to diverse user needs while maintaining a consistent and straightforward presentation format.
3Measurement precision
If a variant-aware ranking model independently ranks each variant item, then the relevance of presented items improves, but the system complexity increases due to multiple scoring and ranking processes
Solution Approach 1:
The patent achieves universality by creating a variant-aware ranking model that handles multiple functions: it scores variants based on query relevance, ranks them independently, and presents them in a unified format. This single multi-functional model resolves the complexity issue by consolidating multiple scoring and ranking processes into one versatile system that improves relevance without proportionally increasing complexity.
4Productivity
If multiple variant items are presented in search results, then user navigation efficiency improves by reducing the need for multiple searches, but the complexity of result presentation increases
Solution Approach 1:
The patent extracts the variant identification and scoring logic from the overall search system, handling it in a separate variant-aware ranking module. This extraction allows the main search interface to remain simple while the extracted component manages the complexity of evaluating and ranking multiple variants, thereby improving user navigation efficiency without significantly increasing the perceived interface complexity.
Data Source
AI summary
Systems and methods of generating interfaces including variant item recommendations are disclosed. A request for an interface and a set of candidate items selected from an item catalog are received. At least one of the candidate items is representative of two or more variant items. A variant score is determined for each variant item related to the at least one of the candidate items and a set of recommended items is generated by independently ranking each item in the set of candidate items and each of the two or more variant items. The set of recommended items is generated by a variant-aware ranking model configured to receive the variant score for each variant item and based on the variant score and a relevancy score for each variant item. The interface including the set of recommended items is generated and transmitted to a system that generated the request for the interface.