Query Error Correction via User Repeat Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems fail to detect errors in speech recognition and natural language understanding, leading to incorrect transcriptions and misinterpretations, which result in negative user experiences.
Innovation Solution
A system and method that systematically identify and correct queries leading to incorrect transcriptions or misinterpretations by using automatic speech recognition and natural language understanding techniques, comparing transcriptions, and analyzing user feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automatic speech recognition systems are used to convert speech to transcriptions, then speech-to-text conversion is achieved, but transcription errors occur due to missed conversion or misinterpretation
Solution Approach 1:
The system implements feedback by detecting when users repeat queries after receiving incorrect responses. The query error detection system monitors user interactions and identifies patterns where users repeat similar queries, indicating transcription or interpretation errors. This feedback loop enables the system to learn from user behavior and correct errors automatically.
Solution Approach 2:
The system performs self-correction by automatically detecting query errors and generating corrected versions without requiring manual intervention. The query error detection system analyzes user repeat queries and automatically generates corrected transcriptions or interpretations, allowing the system to service itself and improve accuracy autonomously.
2Productivity
If natural language understanding systems interpret spoken words, then query understanding is achieved, but misinterpretation errors occur due to incorrect conversion of speech to meaning
Solution Approach 1:
The system uses feedback from user repeat queries to detect misinterpretation errors. When users repeat queries after receiving incorrect responses, the system identifies this pattern as evidence of misinterpretation and uses it to correct future interpretations, improving reliability through observed user behavior feedback.
Solution Approach 2:
The system replaces traditional mechanical speech-to-text conversion with a hybrid approach that incorporates natural language understanding and user feedback analysis. Instead of relying solely on automated transcription, the system uses computational analysis of user interaction patterns to detect and correct interpretation errors.
3Reliability
If the system detects query errors by analyzing user repeat queries, then error detection capability is improved, but system complexity increases due to additional analysis requirements
Solution Approach 1:
The query error detection system leverages existing multi-functional capabilities of the speech recognition and natural language understanding systems. By using the same infrastructure to both process queries and detect errors through repeat query analysis, the system achieves error detection without proportionally increasing complexity, as existing components serve dual purposes.
Data Source
AI summary
Systems and methods are provided for natural language processing using neural network models and natural language virtual assistants. The system and method include receiving a natural language phrase including a word sequence, computing corresponding error probabilities that the words are errors, and for a word with a corresponding error probability above a threshold, then computing a replacement phrase with a low error probability to provide a response from the virtual assistant depending on the replacement phrase.