Update trained voice robots using example-based voice robot development techniques
By updating the corpus of training instances through a voice robot development platform, behavioral errors are automatically identified and corrected, solving the problems of scalability and memory consumption in existing voice robots, and achieving more efficient voice robot performance and comprehension capabilities.
CN115836304BActive Publication Date: 2026-07-17GOOGLE LLC
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GOOGLE LLC
- Filing Date
- 2021-11-22
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Existing voice robots struggle to scale to learn the nuances of human language and require significant computational resources and memory to define and store intent patterns.
Method used
By using a voice robot development platform, the corpus of training instances is updated with a multi-layer machine learning model to automatically identify and correct behavioral errors, and the performance of the voice robot is improved by adding or modifying training instances.
Benefits of technology
It achieves scalability for voice robots, reduces memory consumption, improves accuracy and recall, and enables voice robots to better understand the nuances of human language and respond quickly and efficiently.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN115836304B_ABST
Abstract
The implementation involves updating a trained voice robot deployed to conduct conversations on behalf of a third party. Third-party developers can interact with a voice robot development system that enables them to train, update, validate, and monitor the performance of the trained voice robot. In various implementations, the trained voice robot can be updated by updating the corpus of the training instances initially used to train the voice robot and updating the trained voice robot based on the updated corpus. In some implementations, when conducting conversations on behalf of a third party, the corpus of the training instances can be updated in response to recognizing the occurrence of behavioral errors by the trained voice robot. In additional or alternative implementations, the corpus of the training instances can be updated in response to determining that the trained voice robot does not include the desired behavior.
Need to check novelty before this filing date? Find Prior Art