Variant Item Search Interface Using Multi-Level Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search systems in electronic catalogs often return only a single version of a product, such as the lowest priced variant, regardless of the search query or user intent, requiring additional navigation for users to find matching items.
Innovation Solution
A system and method that utilize a first level search model to generate a baseline value for each item variant as a weighted combination of relevance and engagement label values, and a second level search model to re-rank candidates, generating an interface with variant items that better match user queries and intents, allowing for direct navigation to relevant variant items.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a fixed criteria is used to retrieve and present only a single version of any product from the catalog, then the search system is simple to implement, but the user cannot easily identify an item that matches the search or intent
Solution Approach 1:
The patent segments the search process into two distinct levels: a first level search model that generates a baseline value using a weighted combination of relevance label value and engagement label value, and a second level search model that performs re-ranking. This segmentation allows the system to present multiple variant items rather than a single product version, improving ease of item identification while managing complexity through structured multi-stage processing
Solution Approach 2:
The first level search model performs preliminary action by generating baseline values for multiple item variants before the final presentation. This preliminary ranking based on relevance and engagement metrics prepares a set of candidate items that are then refined by the second level model, allowing users to see multiple relevant variants upfront rather than requiring additional navigation
2Productivity
If the lowest priced variant is returned regardless of search query or intent, then the search system is fast and simple, but additional navigation or interface tasks are required for users to identify a matching item
Solution Approach 1:
The patent changes the ranking parameters from a single fixed criterion (lowest price) to a multi-parameter system that includes relevance label value and engagement label value. The first level search model computes a weighted combination of these parameters, and the second level model further refines ranking, thereby presenting relevant variant items immediately and eliminating the need for additional user navigation
Solution Approach 2:
The engagement label value incorporates feedback from user interactions and behavior patterns to refine search results. By using engagement metrics alongside relevance metrics, the system learns from user feedback to better predict which variant items users are likely to prefer, presenting more relevant results faster without requiring additional navigation time
Data Source
AI summary
Systems and methods for generating an interface including variant item elements are disclosed. A request for an interface including a search query is received from a first device and a set of candidate items is generated by a first level search model based on the search query. The first level search model generates a baseline value for each item variant in an item catalog including a weighted combination of a relevance label value and an engagement label value. The set of candidate items is selected based on the baseline value. A set of search results is generated by a second level search model based on the set of candidate items. An interface including interface elements representative of one or more item variants in the set of search results is generated and transmitted to the first device.