Speech Command Application Selection via Segmented Filtering

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Speech-controlled computing systems face challenges in efficiently determining the appropriate application to execute user commands, leading to increased cognitive burden on users and significant computing resource consumption due to the need to consider multiple potential applications for each command.

Innovation Solution

A system that creates a subset of relevant applications based on the command, allowing for focused natural language processing, reducing the number of applications to be considered and thus minimizing computing resources needed.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the system considers multiple potential applications for each speech command to ensure comprehensive coverage, then the reliability of command execution is improved, but the computing resource consumption and processing time increase significantly

Engineering Contradiction:

Improvecommand execution accuracyVSAvoidprocessing latency

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent segments the application selection process into multiple stages: initial filtering based on command keywords, secondary filtering using application categories, and final selection from a reduced subset. This segmentation allows the system to maintain high reliability by systematically evaluating applications while reducing overall processing time through progressive elimination.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system performs preliminary actions by pre-establishing application categories, keywords, and associations before receiving speech commands. When a command is received, the system quickly retrieves and filters applications based on pre-configured data structures, avoiding the need to analyze all applications from scratch and thus reducing latency while maintaining comprehensive coverage.

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If the system evaluates all available applications for each speech command to ensure accurate application selection, then the measurement precision of command intent is improved, but the device complexity and computing resource consumption increase

Engineering Contradiction:

Improvecommand intent recognition accuracyVSAvoidprocessing system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The evaluation process is segmented into hierarchical levels: first filtering by command keywords, then by application categories, and finally by detailed application features. This segmentation maintains measurement precision by systematically evaluating relevant criteria while reducing device complexity by avoiding unnecessary evaluation of all applications simultaneously.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system applies different levels of evaluation quality to different applications based on their relevance to the command. Highly relevant applications receive detailed evaluation while less relevant ones are quickly filtered out. This local quality approach maintains precision for critical decisions while reducing overall system complexity.

Inventive Principle:

Principle #3Local quality

3Adaptability or versatility

If the system processes natural language commands through multiple applications to ensure comprehensive understanding, then the adaptability of the speech processing system is improved, but the productivity and response speed decrease

Engineering Contradiction:

Improvespeech command handling flexibilityVSAvoidcommand processing throughput

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The system segments the application processing into parallel evaluation threads that work simultaneously on different applications from the filtered subset. This segmentation maintains adaptability by evaluating multiple applications while improving productivity through parallel processing rather than sequential evaluation.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system performs partial evaluation by focusing computational resources on a subset of most relevant applications rather than exhaustively evaluating all applications. This partial action approach maintains sufficient adaptability for accurate command handling while significantly improving processing throughput by avoiding unnecessary evaluation of irrelevant applications.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS11276403B2Natural language speech processing application selection

Publication Date: 2022.03.15 AMAZON TECH INC

AI summary

Techniques for limiting natural language processing performed on input data are described. A system receives input data from a device. The input data corresponds to a command to be executed by the system. The system determines applications likely configured to execute the command. The system performs named entity recognition and intent classification with respect to only the applications likely configured to execute the command.