Query Expansion Using Image Click Vectors and Knowledge Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search engine query expansion methods rely on preset databases and fail to effectively guarantee correlation between queries and expanded queries, lacking richness and diversity in results.
Innovation Solution
A method and apparatus that calculate an image click characteristic vector for a target query, find similar queries in a preset query set based on this vector, and expand entity words and qualifiers using a knowledge graph to determine relevant expanded queries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If query expansion is based on a preset query database with experience-based associations, then the implementation is simple, but the correlation between query and expanded query cannot be effectively guaranteed and the richness and diversity of expanded queries need to be improved
Solution Approach 1:
The patent introduces image click logs as an intermediary data source to bridge the query and expanded query. Instead of directly using experience-based preset databases, the system uses actual user click behavior on images as a mediator to infer query correlations, thereby improving reliability while maintaining implementation feasibility
Solution Approach 2:
The system implements feedback by utilizing user click behavior data from image search results. The click logs provide real-world feedback on which queries lead to relevant images, and this feedback is used to dynamically determine query expansions, ensuring high correlation between queries and their expansions
2Adaptability or versatility
If query expansion uses a preset query database, then the implementation is straightforward, but the richness and diversity of expanded queries are insufficient
Solution Approach 1:
The system enables self-service by automatically generating query expansions from user click behavior data without relying on manual curation of preset databases. The click logs automatically provide the information needed for diverse and relevant query expansions, improving both richness and processing efficiency
Solution Approach 2:
The patent changes the fundamental parameter used for query expansion from static experience-based associations to dynamic user behavior data. By using click frequency and click-through rate as parameters, the system generates diverse and adaptable query expansions that reflect actual user needs
Data Source
AI summary
The present disclosure discloses a method and apparatus for expanding a query. The method comprises: calculating an image click characteristic vector of a target query based on an image click log associated with the target query; finding a similar query of the target query based on the image click characteristic vector, to obtain a candidate expansion query set of the target query; matching the target query and each candidate expansion query in the candidate expansion query set with a knowledge graph, to extract an entity word and a qualifier of the target query and an entity word and a qualifier of the each candidate expansion query; expanding the entity words and the qualifiers of the target query and the each candidate expansion query in combination with the knowledge graph; and matching using expansion results of the entity words and the qualifiers, to determine an expanded query of the target query.