Accessory Search Ranking Using Semantic Graph Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search technologies inaccurately rank and retrieve accessories or sub-accessories, leading to the 'part-of-part' problem, where sub-accessories are ranked higher than the requested accessory, causing unnecessary resource consumption and a poor user experience.
Innovation Solution
Utilizing a word embedding model and knowledge graph to determine semantic relationships between search queries and result candidates, combined with a classifier model, to accurately rank and retrieve accessories or sub-accessories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing search engine ranking technologies are used, then search results are retrieved quickly, but the accuracy of ranking accessories versus sub-accessories deteriorates, causing the part-of-part problem
Solution Approach 1:
The patent segments the search result ranking process into distinct categories: accessories and sub-accessories. By creating separate classification paths and using specific indicators to identify each category, the system accurately distinguishes between them, preventing sub-accessories from incorrectly ranking higher than their parent accessories in search results.
Solution Approach 2:
The patent introduces an intermediary classification mechanism that acts as a mediator between the search query and the ranking process. This intermediary system uses indicators and classification logic to determine whether a search result is an accessory or sub-accessory, thereby mediating the ranking accuracy issue without requiring changes to the underlying search infrastructure.
2Reliability
If existing search technologies retrieve all potential results, then comprehensive results are provided, but resource consumption increases due to redundant queries and processing
Solution Approach 1:
The patent applies preliminary action by classifying and ranking accessories and sub-accessories before the final search result generation. The system pre-processes search results using classification indicators and ranking logic, identifying which results are accessories versus sub-accessories in advance. This preliminary classification prevents redundant processing and reduces the need for follow-up queries, thereby reducing computing resource consumption while maintaining result completeness.
3Measurement precision
If existing search engines rank sub-accessories higher due to keyword matching, then keyword relevance is improved, but the user experience deteriorates due to the part-of-part problem
Solution Approach 1:
The patent applies local quality by implementing different ranking criteria and classification rules for different types of search results. Instead of using a uniform ranking approach, the system applies specific indicators and ranking logic tailored to accessories versus sub-accessories. This localized approach ensures that keyword matching accuracy is maintained for both categories while preventing sub-accessories from incorrectly outranking accessories, thereby improving overall user experience.
Data Source
AI summary
Various embodiments improve search technologies and computer information retrieval by executing a query via ranking a set of search result candidates higher than another set search result candidates based at least in part on the query and determining that a first set of search result candidates are indicative of a sub-accessory to an accessory or an accessory itself.