Dynamic Search Collection Generation for Query-Specific User Intent
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search services fail to accurately reflect users' search intentions by classifying search results into predetermined categories without considering the properties of the query, making it difficult to provide relevant search results.
Innovation Solution
A method and system that automatically generate a search collection for each query by classifying search results into groups based on their subject, using data from user-selected search results, and updating the search collection periodically to reflect changing user interests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If search results are classified according to predetermined search collections, then the search service can provide organized results, but the search results do not reflect user's actual search intention
Solution Approach 1:
The patent transforms static predetermined search collections into dynamic query-specific search collections. The system generates different search collections based on each query's characteristics, making the classification adaptive rather than fixed. This resolves the contradiction by allowing the system to adapt to user search intention without requiring a complex predetermined classification structure for every possible query type.
Solution Approach 2:
The patent changes the parameters of search collection classification from fixed predetermined categories to query-dependent dynamic categories. By analyzing query properties and generating search collections specific to each query, the system adjusts the classification parameters to match user search intention, thereby improving adaptability without increasing overall system complexity.
2Measurement precision
If search collections are generated for each query to reflect user intention, then search result relevance improves, but system complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-processing queries to extract their characteristics and properties before generating search collections. This preparation work is done in advance, allowing the system to efficiently generate appropriate search collections without complex real-time processing, thus improving measurement precision while controlling system complexity.
Solution Approach 2:
The patent uses copying by creating search collections that are specific to each query rather than using a single complex classification system. Each query gets its own tailored search collection based on its properties, which simplifies the overall system architecture while improving precision of search intention reflection.
3Reliability
If search collections are updated periodically to reflect changing user interests, then search result relevance improves, but processing time increases
Solution Approach 1:
The patent implements periodic action by updating search collections at predetermined intervals rather than continuously. This allows the system to maintain up-to-date search collections that reflect changing user interests while avoiding the time loss associated with continuous updating. The periodic updates occur at optimized intervals that balance relevance with processing efficiency.
Data Source
AI summary
A method for generating search collection of query is disclosed, which can provide a search result list displayed by an optimized method of automatically generating a specific collection by each query, the method comprising obtaining a first query and search results selected by a user from a search result list generated in response to the first query; classifying the search results into one or more groups; and generating a search collection for the first query by each group.