Pivot Group Generation for Search Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search and recommendation technologies consume excessive computing resources due to inefficient query processing and item organization, leading to repetitive user inputs and increased storage device I/O.
Innovation Solution
The use of generative models to generate pivot groups that organize search result items or recommended items around themes, facilitating item discovery by assigning items to pivots based on similarity measures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional search and recommendation technologies are used to process queries and organize items, then comprehensive item coverage is achieved, but computing resource consumption increases excessively
Solution Approach 1:
The patent segments the large set of items into multiple pivot groups based on thematic categories. Each pivot group contains items related to a specific theme or aspect, allowing the system to organize and present items in a more structured and efficient manner. This segmentation reduces the computational burden of processing all items uniformly while maintaining comprehensive coverage.
Solution Approach 2:
The patent introduces pivot groups as intermediary structures between the item database and the user interface. These pivot groups act as mediators that organize items thematically, reducing the need for repetitive user queries and decreasing storage device I/O operations. The pivot groups serve as a buffer layer that improves item discovery efficiency while lowering computing resource consumption.
2Ease of operation
If traditional item organization methods are used, then all items can be accessed, but storage device I/O increases due to repetitive user inputs
Solution Approach 1:
The patent performs preliminary organization of items into pivot groups before user interaction. By pre-organizing items into thematic categories, the system reduces the need for users to repeatedly input queries to navigate through unorganized item lists. This preliminary action decreases storage device I/O operations while improving ease of operation.
3Reliability
If comprehensive search processing is performed to ensure all relevant items are found, then item discovery completeness is achieved, but network latency increases
Solution Approach 1:
The patent segments items into pivot groups that can be processed and transmitted more efficiently. By organizing items into thematic categories, the system can retrieve and transmit relevant pivot groups faster than processing individual items, reducing network latency while maintaining item discovery completeness.
Data Source
AI summary
Some aspects relate to technologies for generating pivots using a generative model and grouping items into pivot groups using the pivots. Pivots are generated by obtaining item information for items, generating a prompt using the item information, and causing a generative model to use the prompt to generate text for the pivots, including a pivot name and pivot description for each pivot. A pivot embedding is also generated for each pivot. When items are to be returned to a user device (e.g., as recommended items or search result items), item embeddings for the items and the pivot embeddings for the pivots are used to assign each item to a particular pivot to generate pivot groups that each includes a pivot name, a pivot description, and assigned items. User interface information is provided to present at least a portion of the pivot groups on a user device.