Query Resolution System Integrating Structured and Unstructured Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Curated knowledge bases are limited in providing answers to queries outside their specific domain, as they cannot resolve subqueries that involve information missing from their structured data, leading to incomplete information and reduced accuracy in question answering systems.
Innovation Solution
A method that receives a natural language question, derives subqueries, determines which cannot be resolved using a curated knowledge base, resolves them using unstructured text, and updates the knowledge base with the resolved information, integrating structured and unstructured data to enhance the knowledge base's coverage and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a curated knowledge base is used to provide accurate answers, then measurement precision is improved, but adaptability or versatility deteriorates because the knowledge base cannot resolve queries outside its specific domain
Solution Approach 1:
The system merges a curated knowledge base with unstructured text data sources to create a hybrid query resolution approach. The knowledge base provides structured, high-precision answers for domain-specific queries while unstructured text supplements the system with broader knowledge coverage, resolving the contradiction between precision and versatility.
Solution Approach 2:
An intermediary layer is introduced between the user query and the knowledge base. This layer includes query decomposition into subqueries, determination of resolvable vs. unresolvable subqueries, and integration of unstructured text results, enabling the system to handle queries outside the knowledge base domain while maintaining accuracy for domain-specific queries.
2Manufacturing precision
If a curated knowledge base is used to provide high-precision knowledge, then manufacturing precision is improved, but device complexity increases when handling queries outside the specific subset
Solution Approach 1:
The query resolution process is segmented into distinct stages: receiving the natural language question, deriving subqueries, determining which subqueries can be resolved by the knowledge base and which require unstructured text, and integrating results. This segmentation manages complexity by breaking down the overall system into manageable functional components.
Solution Approach 2:
The system dynamically adapts its behavior based on the query type. For queries within the knowledge base domain, it uses structured data for fast, precise responses. For queries outside the domain, it dynamically switches to unstructured text search. This dynamic approach maintains precision for relevant queries while avoiding the complexity of processing all queries through a single rigid system.
3Adaptability or versatility
If unstructured text is used to resolve subqueries, then adaptability or versatility is improved, but measurement precision deteriorates due to lower quality of unstructured data
Solution Approach 1:
The system uses feedback from the query resolution process to improve future performance. By analyzing which subqueries require unstructured text and why, the system can refine its query decomposition and data selection strategies, gradually improving the quality of answers from unstructured sources while maintaining versatility.
Solution Approach 2:
The system applies different quality standards and processing methods to different parts of the query. For subqueries resolvable by the knowledge base, it uses high-precision structured data. For subqueries requiring unstructured text, it applies specialized processing to maximize quality. This localized approach optimizes overall system performance by not requiring uniform quality across all data sources.
Data Source
AI summary
A computer-implemented method according to one embodiment includes receiving a question in a natural language format; deriving a plurality of subqueries from the question; determining that a first subquery of the plurality of subqueries cannot be resolved utilizing a curated knowledge base; resolving the first subquery, utilizing unstructured text; and adjusting the curated knowledge base, utilizing the resolved first subquery.