Sub-question Result Merging in QA Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Question and Answer (QA) systems face difficulties in accurately answering overly broad or differently phrased input questions due to the presence of pragmatic implications in reference documents, leading to information gaps and reduced accuracy.
Innovation Solution
The implementation of a mechanism that identifies and addresses sub-questions related to the input question, allowing the QA system to operate on each sub-question separately and merge results to provide a consolidated answer, employing domain adaptation techniques and natural language processing to map broad references to specific questions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the QA system operates on broadly phrased input questions directly, then the system simplicity is maintained, but the answer accuracy deteriorates due to pragmatic implications and information gaps in reference documents
Solution Approach 1:
The patent applies segmentation by dividing a broad input question into multiple specific sub-questions. Each sub-question targets a particular aspect of the original question, allowing the QA system to retrieve more precise and relevant information from reference documents. The results from multiple sub-questions are then merged to form a comprehensive answer, thereby improving answer accuracy while managing system complexity through structured decomposition.
2Loss of information
If the system merges results from multiple sub-questions, then the information completeness is improved, but the processing time increases
Solution Approach 1:
The patent applies preliminary action by pre-identifying and decomposing the input question into specific sub-questions before information retrieval. This upfront structuring allows the system to target its information gathering more efficiently, reducing redundant searches and enabling better parallel processing of sub-questions. The preliminary decomposition thus improves information completeness while mitigating processing time increases through more focused and efficient data collection.
3Measurement precision
If domain adaptation techniques are applied to map broad references to specific questions, then the relevance of retrieved information is improved, but the computational complexity increases
Solution Approach 1:
The patent applies the intermediary principle by introducing sub-questions as intermediate elements between the broad input question and the reference documents. These sub-questions act as mediators that translate vague, broadly phrased questions into specific, targeted information retrieval queries. This intermediary layer improves information relevance by ensuring that retrieved documents directly address specific aspects of the original question, while managing computational complexity through structured, modular processing.
Data Source
AI summary
An input question enhancement mechanism is provided for merging sub-question results in Question and Answer (QA) systems in order to answer an input question submitted from a user. A plurality of sub-questions associated with the input question is identified. Responsive to submitting the set of questions to a QA pipeline of the QA system, a set of answers and a set of passages related to the set of answers corresponding to each of the plurality of sub-questions are received. A mapping is applied to the set of passages to form a consolidated set of passages. The set of answers is then merged into the mapping so as to generate a consolidated set of answers. A factoid question response report is generated with answers from the consolidated set of answers and passages from the consolidated set of passages. The factoid question response report is then provided to the user.