Deep Question Answering System Handling Unspecified Input Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Deep question answering systems generate low-confidence answers when users fail to specify critical information in their queries, leading to less useful responses.
Innovation Solution
Implementing a parallel processing approach within the deep QA system to evaluate questions with multiple alternative values for unspecified critical data, allowing the system to generate candidate answers with higher confidence scores without re-processing the original question.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the deep QA system generates candidate answers without requiring critical information specification, then the system can answer more questions, but the confidence scores of the answers decrease
Solution Approach 1:
The system performs preliminary action by proactively identifying missing critical information elements before generating final answers. It generates multiple candidate answers with different imputed values for the missing information, allowing the system to maintain high confidence scores by selecting answers supported by strong evidence from the corpus, rather than blindly generating answers without verification.
2Measurement precision
If the deep QA system processes questions with multiple alternative values for unspecified critical data, then the accuracy of answers improves, but the processing complexity increases
Solution Approach 1:
The system segments the answer generation process into distinct paths based on different imputed values for missing critical information. It processes each candidate answer path independently, evaluating evidence from the corpus for each scenario. This segmentation allows the system to manage complexity by handling one alternative at a time while maintaining overall accuracy through comprehensive evaluation of multiple possibilities.
3Adaptability or versatility
If the deep QA system uses imputed values for missing critical information, then more complete answers can be generated, but the risk of generating inaccurate answers increases
Solution Approach 1:
The system implements feedback mechanisms by evaluating generated candidate answers against evidence from the corpus. It scores and ranks candidate answers based on their support from the corpus, allowing the system to identify and discard inaccurate imputed values. This feedback loop ensures that only answers with sufficient evidential support are returned, maintaining reliability while enabling comprehensive answer generation.
Data Source
AI summary
System, method, and computer program product for performing an operation, the operation comprising receiving, by a deep question answering system, a question not specifying an element of input data, identifying a set of possible values for the element of input data, generating, by the deep question answering system, a respective set of candidate answers for the question: (i) without the element of input data, and (ii) with each possible value in the set of possible values for the element of input data, and returning at least one candidate answer from at least one set of candidate answers as responsive to the question.