Query Expansion Search for Diverse Item Discovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search systems on E-commerce sites struggle to display diverse and relevant items related to user queries due to insufficient data on item relationships, limiting the discovery of new and varied items.
Innovation Solution
An information processing system that acquires a first query, extracts items and associated queries, and outputs information based on both queries to enhance item discovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If item relationship data is used to find related items, then relevant items can be displayed, but the data is not always sufficient leading to reduced variety
Solution Approach 1:
The patent segments the item discovery process into multiple independent pathways: (1) using pre-stored item relationship data to find related items, (2) using query expansion to generate alternative search queries, and (3) using search history to find items. Each pathway operates independently and contributes to the final diverse set of displayed items, resolving the contradiction between relying on relationship data and achieving variety.
Solution Approach 2:
The patent merges multiple item discovery methods by combining results from item relationship data, query expansion, and search history into a unified item set. This integration allows the system to leverage the reliability of relationship data while simultaneously incorporating diverse items from alternative queries and historical searches, thus achieving both accuracy and variety.
2Speed
If only simple query search is performed, then search speed is fast, but the items found are not desirable enough for users
Solution Approach 1:
The patent performs preliminary actions by pre-storing item relationship data and maintaining search histories before the actual search occurs. During search execution, these pre-prepared resources enable rapid retrieval of related items and alternative queries without adding significant processing time, thus maintaining search speed while improving item desirability through expanded results.
Solution Approach 2:
The patent introduces query expansion as an intermediary mechanism that bridges the simple user query and the final item results. By generating alternative queries from the original query and using them to search for additional items, the system enhances item desirability without requiring complex real-time processing, thus preserving search speed.
3Adaptability or versatility
If multiple item discovery methods are used, then item variety increases, but system complexity increases
Solution Approach 1:
The patent segments the complex item discovery system into three independent, modular components: item relationship data retrieval, query expansion generation, and search history utilization. Each component operates independently with well-defined interfaces, making the overall system manageable despite handling multiple discovery methods. This modular segmentation reduces system complexity while maintaining item variety.
Solution Approach 2:
The patent creates a universal item discovery framework where a single system architecture supports multiple discovery methods (relationship-based, query-expansion-based, and history-based). This multi-functional design allows the system to achieve high item variety through a unified structure rather than separate complex systems, thereby managing overall system complexity.
Data Source
AI summary
An information processing system acquires a first query; extracts at least one item from a plurality of items stored in advance, based on the first query; acquires another item associated with the at least one item being extracted; acquires at least one query associated with the other item as a second query; and outputs information according to the second query together with an item retrieved based on the first query.