Query Clustering for Search Result Ranking

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Online retailers face challenges in providing relevant search results due to the large number of products available, leading to many irrelevant matches in free-form text searches.

Innovation Solution

A system that clusters user queries, assigns similarity scores, and uses these clusters to enhance search results by associating categories and identifying synonyms, thereby improving the relevance of search outcomes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If free-form text search is used to search large product databases, then the search coverage is broad, but the search result relevance deteriorates due to many irrelevant matches

Engineering Contradiction:

Improvesearch coverageVSAvoidsearch result relevance

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent segments the search process into multiple stages: query clustering to group similar queries, category association to identify product categories, and synonym expansion to capture related terms. This segmentation allows the system to handle broad search coverage while improving relevance through structured processing of search queries and results.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary processing layer that includes query clustering modules, category association modules, and synonym expansion modules. These intermediaries process the raw search queries and transform them into enhanced search parameters, filtering out irrelevant results while maintaining comprehensive coverage.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If the number of products available is increased to meet diverse customer needs, then the product variety is improved, but the search difficulty increases due to more irrelevant products

Engineering Contradiction:

Improveproduct varietyVSAvoidsearch difficulty

Core Design Contradiction:

Adaptability or versatilityVSEase of operation

Solution Approach 1:

The patent performs preliminary actions by pre-clustering queries into groups, pre-associating categories with query clusters, and pre-expanding synonyms before the actual search execution. This preliminary processing reduces the search difficulty by preparing organized search parameters in advance, making it easier to navigate large product catalogs with high variety.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent incorporates feedback mechanisms where search results are analyzed to refine query clustering and category associations. User interactions and search patterns are used to continuously improve the relevance modeling, making the search easier to operate as the system learns from actual usage patterns.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS9251292B2Search result ranking using query clustering

Publication Date: 2016.02.02 WALMART APOLLO LLC

AI summary

Systems and methods are disclosed herein for providing search results, particularly product records from a product database. Past queries are analyzed and grouped into clusters according to similarities, with each query including a highly relevant head query and a plurality of tail queries. Similarity of queries may be determined based on user response similarity to query results, co-occurrence, and textual similarity. One or more categories are identified for the clusters, such as based on click-through rates for search results of the head queries of the clusters. Upon receiving a query, a cluster for the query is identified, such as according to similarity to one or more queries of the cluster. The categories associated with the cluster are then used to one or both of augment the query and rank search results for the query.