Visual Diversity Ranking for Real-Time Search Results
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search result ranking techniques often cluster identical or similar items together, leading to a suboptimal user experience in large-scale databases, making it difficult for users to identify desired items.
Innovation Solution
Implement a visual diversity ranking function that utilizes image hashes to reorder search results based on visual similarities, ensuring a balanced mix of relevant and diverse items is presented.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If search results are ranked solely by relevance, then the most relevant items are presented first, but identical or similar items are clustered together making it difficult for users to identify desired items
Solution Approach 1:
The patent segments the search result ranking process into two distinct phases: first ranking by relevance, then applying visual diversity reordering. This segmentation allows the system to maintain relevance accuracy while preventing clustering of similar items through the second diversity-based ordering pass.
Solution Approach 2:
The patent creates a composite ranking approach that combines relevance scoring with visual diversity metrics. By integrating these two different criteria (relevance and visual diversity) into a unified ranking system, the patent achieves both accurate relevance presentation and improved user experience through diverse visual representation.
2Quantity of substance
If multiple identical or similar items are presented consecutively, then all relevant items are shown to users, but user experience is frustrated and item identification becomes difficult
Solution Approach 1:
The patent introduces dynamic reordering of search results based on visual diversity metrics. Instead of a static relevance-only ranking, the system dynamically adjusts the presentation order to interleave similar items, maintaining the complete set of relevant results while improving their visual distribution and identifiability.
3Ease of operation
If a visual diversity ranking function is implemented, then visual diversity of search results is improved, but system complexity increases
Solution Approach 1:
The patent performs preliminary action by pre-computing visual features and similarity metrics for items before the actual search ranking occurs. This pre-processing allows the visual diversity ranking function to operate efficiently during query execution, reducing the computational burden and system complexity during real-time search operations.
Data Source
AI summary
A system including one or more processors and one or more non-transitory computer-readable storage devices storing computing instructions that, when run on the one or more processors, cause the one or more processors to execute operations: storing image hashes corresponding to images for items offered via an electronic platform; receiving, from a computing device, a search query via a search engine included on the electronic platform; executing a relevance function; executing a visual diversity ranking function that reorders the relevant set of search results to generate a ranked set of search results, wherein the visual diversity ranking function utilizes the image hashes corresponding to the subset of the items to reorder the relevant set of search results; and transmitting the ranked set of search results to the computing device. Other embodiments are disclosed herein.