Shared Custom Grammars for Scalable Voice Command Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for multi-channel interaction applications face challenges in efficiently processing user utterances across various communication channels, particularly in recognizing complex spoken commands, which leads to increased development costs and resource utilization due to the need for custom grammars that are not easily scalable or shareable among different interaction sites.
Innovation Solution
A communications system that receives requests to execute interaction sites associated with custom grammars files, updates these files by storing and processing user utterances, and generates private custom grammars files accessible only to specific interaction sites, enabling the recognition of new commands and sharing of updated grammars across multiple sites.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If custom grammars are developed for each interaction site independently, then each site can recognize specific commands accurately, but development costs and resource utilization increase significantly
Solution Approach 1:
The patent implements a shared custom grammars repository that multiple interaction sites can access and utilize. Instead of each site maintaining separate grammar files, the system allows grammars to be developed once and reused across multiple sites, reducing development complexity while maintaining command recognition accuracy through the shared repository mechanism.
Solution Approach 2:
The system merges previously separate grammar development efforts into a unified shared repository. Multiple interaction sites are combined into a single ecosystem that collaboratively uses the same grammar resources, eliminating redundant development work while preserving the ability to recognize site-specific commands through selective grammar assignment.
2Measurement precision
If custom grammars are developed for each interaction site independently, then each site has dedicated command recognition capability, but the system becomes less scalable and harder to maintain
Solution Approach 1:
The shared custom grammars repository provides universal access to command recognition resources across multiple interaction sites. This universal mechanism allows new sites to be added to the system without requiring independent grammar development, thereby improving scalability while maintaining the precision of command recognition through the shared grammar library.
Solution Approach 2:
The system implements dynamic grammar sharing where grammars can be selectively assigned to different interaction sites based on their needs. This dynamic allocation allows the system to adapt to new sites and requirements flexibly, improving scalability while maintaining precise command recognition by assigning only the necessary grammars to each site.
3Device complexity
If a shared custom grammars repository is implemented across multiple interaction sites, then development costs are reduced and scalability improves, but managing and updating grammars across multiple sites becomes more complex
Solution Approach 1:
The system implements a centralized grammar repository with version control and update propagation mechanisms. When grammars are updated in the shared repository, the system provides feedback to all interacting sites, automatically or selectively applying updates across the ecosystem. This feedback mechanism simplifies grammar management by centralizing control while maintaining consistency across multiple sites.
Solution Approach 2:
The shared custom grammars repository acts as an intermediary between grammar developers and interaction sites. This intermediary layer manages the complexity of grammar storage, versioning, and distribution, shielding individual sites from the complexity of manual grammar management while enabling efficient updates and reductions in development complexity.
4Measurement precision
If grammars are updated frequently to recognize new utterances, then command recognition capability improves over time, but system resource utilization increases
Solution Approach 1:
The system performs preliminary processing of user utterances by collecting and storing them in the shared grammar repository before formal grammar generation. This preliminary action allows the system to accumulate sufficient data and patterns before committing resources to full grammar compilation and distribution, thereby improving command recognition accuracy over time while managing processing resource utilization through staged processing.
Solution Approach 2:
The system implements selective grammar updates where only the necessary portions of grammars are regenerated and distributed when new utterances are learned. Instead of fully recompiling all grammars, the system performs partial updates targeted at specific grammar rules that need modification, improving command recognition capability while reducing the overall processing resource utilization compared to complete grammar regeneration.
Data Source
AI summary
A request to execute an interaction site associated with a custom grammars file is received from a user device and by a communications system. An interaction flow document to execute the interaction site is accessed by the communications system. The custom grammars file is accessed by the communications system, the custom grammars file being configured to enable the communications system to identify executable commands corresponding to utterances spoken by users of user devices. An utterance spoken by a user of the user device is received from the user device and by the communications system. The utterance is stored by the communications system. The custom grammars file is updated by a grammar generation system to include a representation of the stored utterance for processing utterances in subsequent communications with users.