AI Answer Merging for Contextual Feature Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current question answering systems often provide inaccurate answers by selecting irrelevant passages, as they prioritize single answers over comprehensive passages that better match contextual features, leading to suboptimal responses.
Innovation Solution
A method is developed to train an AI system to identify contextual features in questions, rank passages based on feature matching, merge multiple relevant passages into a single answer, and replace initial answers with merged answers that have a higher total quantity of contextual features, thereby improving answer accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the AI system selects the single highest-ranked passage as the answer, then the answer generation is simple and fast, but the answer accuracy and comprehensiveness deteriorate
Solution Approach 1:
The patent merges multiple candidate passages into a single comprehensive answer. The system identifies the highest-ranked passage and additional lower-ranked passages, then combines their content through string concatenation to create a merged answer that incorporates contextual features from multiple sources, thereby improving answer accuracy while managing complexity through automated processing
2Loss of information
If the AI system merges multiple passages into a single answer, then the comprehensiveness and contextual feature coverage improve, but the processing time and computational resources increase
Solution Approach 1:
The patent applies partial merging by selectively combining only the highest-ranked passage with a limited number of additional lower-ranked passages, rather than merging all possible candidate passages. This approach achieves sufficient contextual feature coverage without the excessive processing time that would result from comprehensive merging of all candidates
3Reliability
If the AI system prioritizes single answer selection, then the system complexity remains low, but the answer relevance and contextual matching deteriorate
Solution Approach 1:
The patent performs preliminary identification and ranking of multiple candidate passages before final answer selection. The system pre-processes candidate passages by identifying contextual features, ranking them by relevance, and selecting a subset for merging, thereby ensuring answer relevance while managing system complexity through structured preliminary processing
Data Source
AI summary
A method trains and utilizes an artificial intelligence (AI) system. The AI system receives a question that has contextual features. The method trains the AI system to identify entries in a corpus that have one or more of the contextual features from the question. The method further trains the AI system to: form a set of answers to the question based on identified contextual entries in the corpus; identify and name an entry in the corpus that has a highest quantity of contextual features that match the contextual features in the question as an initial answer to the question; identify and merge multiple other answers to the question from the corpus; and replace the initial answer with the merged answer in order to create a fully trained AI system. The fully trained AI system is then utilized to answer the question with the merged answer.