Clustering Search Results Using Entity Ontology
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional search engines provide search results in a ranked list that lacks logical structure, leading to dominance by popular items and insufficient diversity, making it difficult for users to explore search items effectively.
Innovation Solution
A system that clusters search results using entities in a knowledge base, forming clusters based on entity ontology relationships, and evaluates them using metrics like coverage, balance, overlap, and silhouette ratio to provide coherent and diverse search results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If search results are provided as a conventional ranked list, then popular items can be easily identified and accessed, but the results lack logical structure and diversity, making it difficult for users to explore different categories of items
Solution Approach 1:
The patent segments the flat ranked list of search results into hierarchical clusters based on entity ontology relationships. Each cluster groups items sharing common ontological ancestors, creating a tree-like structure that preserves logical relationships while maintaining ease of access to popular items within each cluster.
Solution Approach 2:
The patent adds a hierarchical dimension to the traditional linear ranked list by organizing results into multiple levels of clustering based on ontology depth. This transforms the one-dimensional ranked list into a multi-dimensional hierarchical structure that simultaneously provides overview and detailed exploration capabilities.
2Loss of information
If search results are clustered using entity ontology, then logical structure and diversity are improved, but the system complexity and computational requirements increase
Solution Approach 1:
The patent performs preliminary actions by pre-computing entity ontologies and item-to-entity mappings before search execution. This allows the search system to leverage pre-built hierarchical structures during query processing, reducing real-time computational complexity while maintaining rich logical structure in results.
Solution Approach 2:
The patent introduces entity ontologies as an intermediary layer between search queries and result items. This mediator structure enables complex hierarchical organization without requiring direct complex processing between queries and items, simplifying the overall system architecture while preserving logical relationships.
3Manufacturing precision
If multiple clustering methods are evaluated in parallel, then the quality of clusters is improved through comprehensive evaluation, but the computational time and resources increase
Solution Approach 1:
The patent applies partial evaluation by computing cluster quality metrics for only the most promising clustering methods or configurations identified through preliminary filtering. This selective approach maintains high cluster quality by evaluating multiple methods while avoiding the full computational cost of exhaustively evaluating all possible clustering approaches.
Data Source
AI summary
Implementations provide an improved system for presenting search results based on entity associations of the search items. An example method includes generating first-level clusters of items responsive to a query, each cluster representing an entity in a knowledge base and including items mapped to the entity, merging the first-level clusters based on entity ontology relationships, applying hierarchical clustering to the merged clusters, producing final clusters, and initiating display of the items according to the final clusters. Another example method includes generating first-level clusters from items responsive to a query, each cluster representing an entity in a knowledge base and including items mapped to the entity, producing final clusters by merging the first-level clusters based on an entity ontology and an embedding space that is generated from an embedding model that uses the mapping, and initiating display of the items responsive to the query according to the final clusters.