Query System Management via Feedback-Based Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current query systems require users to submit questions to multiple systems, leading to complex operations due to varying answers from different knowledge databases, necessitating a more efficient method for managing and selecting the appropriate query system based on user feedback.
Innovation Solution
A computer-implemented method and system that determines an association relationship between questions and query systems using user feedback, allowing for the selection of the most suitable query system for answering similar questions, thereby improving answer accuracy and user satisfaction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If users submit questions to multiple query systems, then the system can provide comprehensive answers from different knowledge databases, but the operation complexity increases and user experience deteriorates
Solution Approach 1:
The patent introduces a query management system as an intermediary layer between the user and multiple query systems. This mediator automatically routes questions to appropriate query systems based on learned associations, eliminating the need for users to manually submit queries to multiple systems while still achieving comprehensive answer coverage through the backend coordination of multiple knowledge databases.
2Reliability
If multiple query systems are used to answer questions, then answer diversity and completeness improve, but the difficulty of selecting the appropriate system increases
Solution Approach 1:
The patent implements a feedback mechanism where user interactions (clicks, satisfaction ratings, correction actions) are continuously collected and used to update the association relationships between questions and query systems. This feedback loop enables the system to learn from actual user behavior and automatically optimize its selection strategy, making the system selection process transparent and effective without requiring user intervention.
Solution Approach 2:
The system performs preliminary learning and association building in advance by analyzing historical query data and user feedback patterns. This preliminary action creates a pre-established mapping between question types and suitable query systems, allowing the system to quickly and accurately select the appropriate query system when a new question is received, without requiring complex real-time analysis.
3Measurement precision
If query systems are selected based on user feedback, then answer accuracy and user satisfaction improve, but the time required to process and learn from feedback increases
Solution Approach 1:
The patent applies partial action by selectively updating association relationships based on significant user feedback events rather than processing every interaction equally. The system focuses on meaningful feedback (such as explicit corrections or strong satisfaction signals) to update its models, avoiding unnecessary processing of routine interactions. This approach maintains high answer accuracy while reducing the overall feedback processing time and computational overhead.
Data Source
AI summary
Embodiments of the present invention relate to a method, system, and computer program product for query system management. In a method, a question that is to be processed by a group of query systems is received. A group of answers to the question are received from the group of query systems, wherein the group of answers comprise an answer to the question from each of the query systems associated with the group of query systems. An association relationship is determined between the question and a query system from the group of query systems, wherein the association relationship is determined based on user feedback to the answer that is provided by the query system, and wherein the association relationship includes an association between the query and the query system and is used for selecting the query system for answering the question and further questions that are determined to be similar.