NLU Intent Adjustment for Subscription-Limited Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Speech processing systems face challenges in executing user intents on devices like vehicles without proper authorization, often requiring data subscriptions, leading to limitations in user interaction and experience.
Innovation Solution
The system determines a secondary NLU intent that can be executed locally by the device, bypassing authorization requirements, allowing users to interact with speech processing systems without needing a data subscription, by converting requests into alternative actions that can be performed using device-executable intents.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the system requires data subscription for executing NLU intents on devices, then authorization and control are improved, but user interaction freedom and system accessibility deteriorate
Solution Approach 1:
The patent segments NLU intents into two categories: cloud-dependent intents requiring data subscription and device-executable intents that can run locally. This segmentation allows the system to maintain authorization control for cloud services while enabling free interaction for local capabilities, resolving the contradiction between reliability and adaptability
Solution Approach 2:
The patent introduces an intermediary component that translates user requests into appropriate intent types. When a user interacts with the system, the intermediary determines whether the request can be handled by device-executable intents or requires cloud processing, thereby mediating between user freedom and authorization requirements
2Adaptability or versatility
If the system translates user requests into device-executable intents, then user interaction freedom is improved, but system functionality and service quality may deteriorate
Solution Approach 1:
The patent implements a dynamic intent translation mechanism that adapts to user needs and system capabilities. The system dynamically determines whether to translate requests into device-executable intents or maintain them as cloud-dependent intents based on available resources, user preferences, and request complexity, balancing freedom and quality
Solution Approach 2:
The patent changes the parameter of intent executability by introducing a spectrum from fully cloud-dependent to fully device-executable. By adjusting this parameter based on specific request characteristics, the system can optimize the balance between user interaction freedom and service quality for different scenarios
3Power
If cloud-based speech processing is used, then processing capability is improved, but data transmission requirements and operational limitations worsen
Solution Approach 1:
The patent segments processing capabilities between cloud and device, with complex NLU processing in the cloud and simpler intent execution on the device. This segmentation reduces data transmission requirements by only sending necessary requests to the cloud while handling local execution independently
Solution Approach 2:
The patent extracts the essential intent recognition functionality from cloud-based processing and implements it locally on devices. This extraction allows the system to maintain high processing capability for critical functions while eliminating the need for continuous cloud connectivity and data transmission
Data Source
AI summary
Techniques for responding to a user input using a device-executable intent are described. A user may speak, to a device, an input corresponding to a first natural language understanding (NLU) intent executable by a speech processing system. The speech processing system may determine a device type corresponding to the device and, based on the device type, determine a data subscription is required to execute the first NLU intent. Thereafter, the speech processing system may determine a second NLU intent that is executable by the device and that results in the output of similar content to the execution of the first NLU intent. The speech processing system then directs the device to execute the second NLU intent.