Multi-command Voice Input Parsing for Virtual Assistants
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face inefficiencies and frustration when interacting with virtual assistants, as they can only address one task at a time through speech, leading to delays and broken concentration, making it time-consuming to manage multiple tasks or notifications.
Innovation Solution
A system processes multi-part voice commands by transcribing speech input into text, parsing it into candidate substrings, determining actionable commands based on probability, and executing associated user intents with acknowledgments, including audible or haptic feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a virtual assistant processes only one task at a time through speech input, then the system complexity remains manageable, but the productivity and user efficiency deteriorate due to time-consuming sequential task handling
Solution Approach 1:
The system segments the speech input into multiple candidate substrings, each representing a potential actionable command. This segmentation allows the virtual assistant to process multiple tasks simultaneously from a single utterance, improving productivity without overwhelming the system with undivided complex input
Solution Approach 2:
The system performs partial processing by generating multiple candidate substrings with associated probabilities, then selectively processes only those exceeding a threshold. This approach handles multiple tasks efficiently while maintaining manageable complexity by filtering out low-probability candidates
2Reliability
If a user must wait for each virtual assistant task to complete before moving to the next task, then the system ensures reliable task execution, but the loss of time increases due to sequential processing delays
Solution Approach 1:
The system performs preliminary parsing of the speech input into multiple candidate substrings and determines their probabilities before full task execution. This allows the system to identify and prepare multiple actionable commands in advance, enabling parallel processing that reduces total time while maintaining reliability through pre-validation
Solution Approach 2:
The system enables continuous processing by handling multiple tasks from a single utterance without requiring waiting periods between tasks. By processing all high-probability candidate substrings simultaneously, the system eliminates idle time while maintaining execution reliability through comprehensive intent determination
3Productivity
If the system processes multiple actionable commands from a single utterance, then the productivity improves through parallel task handling, but the difficulty of detecting and measuring increases due to parsing complexity
Solution Approach 1:
The system uses probability scores as feedback to evaluate each candidate substring's likelihood of being an actionable command. This quantitative feedback mechanism simplifies the detection process by providing clear metrics for filtering and selection, making it easier to identify valid commands among multiple candidates
Solution Approach 2:
The system changes the parameter of command detection from binary recognition to probabilistic scoring. By assigning probability values to each candidate substring and applying a threshold filter, the system transforms a complex detection problem into a manageable parameter-based filtering process
4Ease of operation
If the system provides detailed acknowledgments for multiple user intents, then the ease of operation improves through clear user feedback, but the device complexity increases due to multiple feedback mechanisms
Solution Approach 1:
The system merges multiple acknowledgment mechanisms (audible confirmation, haptic feedback, visual display) into a unified feedback response for processing multiple user intents. This consolidated approach provides comprehensive user feedback without requiring separate complex feedback systems for each task
Data Source
AI summary
Systems and processes are disclosed for handling a multi-part voice command for a virtual assistant. Speech input can be received from a user that includes multiple actionable commands within a single utterance. A text string can be generated from the speech input using a speech transcription process. The text string can be parsed into multiple candidate substrings based on domain keywords, imperative verbs, predetermined substring lengths, or the like. For each candidate substring, a probability can be determined indicating whether the candidate substring corresponds to an actionable command. Such probabilities can be determined based on semantic coherence, similarity to user request templates, querying services to determine manageability, or the like. If the probabilities exceed a threshold, the user intent of each substring can be determined, processes associated with the user intents can be executed, and an acknowledgment can be provided to the user.