Contextual Query Suggestions via Knowledge Graph and Passage Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional search engines provide less relevant query suggestions based on user-submitted queries and typically present them at the top or bottom of search engine results pages (SERPs), which may not align with the user's current interests, and require extensive searches across multiple sources to supplement passage information.
Innovation Solution
A computing system generates suggested queries using a generative model, such as a transformer model, that combines knowledge graph data with passage information from SERPs, allowing for contextually relevant query suggestions to be presented when a user hovers over a passage, eliminating the need for previously submitted queries and reducing resource usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional search engines provide query suggestions based on user-submitted queries, then the suggestions can be generated, but the relevance to user's current interests deteriorates and extensive searches across multiple sources are required
Solution Approach 1:
The system performs preliminary action by pre-processing passage information and knowledge graph data to generate contextual query suggestions before the user actually needs them. The generative model continuously learns from passage content and knowledge graphs, preparing relevant suggestions in advance so they can be immediately presented when user interaction occurs, eliminating the need for extensive real-time searches.
Solution Approach 2:
The knowledge graph serves as an intermediary that bridges passage information and query suggestions. Instead of directly analyzing passages and generating suggestions, the system uses the knowledge graph as an intermediate layer that structures entity relationships and attributes, enabling more efficient and relevant suggestion generation without requiring extensive searches across multiple sources.
2Ease of operation
If suggested queries are presented at the top or bottom of SERP, then they can be displayed, but they do not align with user's current interests and require extensive searches
Solution Approach 1:
The system applies local quality by presenting query suggestions in a pop-up graphical element specifically located proximate to the passage the user is viewing, rather than uniformly at the top or bottom of the SERP. This localized placement ensures suggestions are contextually relevant to the specific passage content, improving alignment with user interests while reducing the time needed to find relevant information.
Solution Approach 2:
The suggestion presentation is dynamic rather than static. The pop-up graphical element appears interactively when the user selects or hovers over a passage, adapting the suggestion display to the user's real-time interaction pattern. This dynamic presentation ensures suggestions are shown exactly when and where they are most relevant, eliminating the need for extensive searches.
3Loss of information
If a generative model combines knowledge graph data with passage information, then contextually relevant query suggestions are generated, but computational resources are consumed
Solution Approach 1:
The knowledge graph serves multiple functions simultaneously: it structures entity information, establishes relationships between entities and attributes, and provides contextual data for query suggestion generation. By making the knowledge graph multi-functional, the system reduces the need for separate processing steps and minimizes overall computational resource consumption while maintaining high contextual relevance in suggestions.
Solution Approach 2:
The system optimizes computational resource usage by dynamically adjusting processing parameters based on the complexity of the passage and the depth of knowledge graph queries required. The generative model can adapt its processing intensity to match the actual information needs, generating contextually relevant suggestions without consistently consuming maximum computational resources.
4Loss of information
If suggested queries are presented in a pop-up graphical element proximate to the passage, then relevance is improved, but device complexity increases
Solution Approach 1:
The pop-up graphical element is nested within the existing SERP structure, integrating seamlessly with the current web page layout. The suggestion element is contained within or adjacent to the passage element, creating a nested hierarchical presentation that adds relevance without significantly increasing overall device complexity or disrupting the existing user interface architecture.
Data Source
AI summary
A computing system receives an indication that a user has selected a passage shown on a search engine results page (SERP) presented on a display. Upon receiving the indication, the computing system identifies a plurality of suggested queries related to the passage, where the plurality of suggested queries are generated based upon the passage and an entry for an entity in a knowledge graph. Upon identifying the plurality of suggested queries, the computing system presents the plurality of suggested queries in a pop-up graphical element that overlays a portion of the SERP, where the pop-up graphical element is located proximate to the passage shown on the SERP. When a query in the plurality of suggested queries is selected, a second SERP is presented on the display, where the second SERP is based upon the query.