Obstacle removing intelligent agent training method and device, and electronic equipment
By validating the format and content of the troubleshooting process corpus samples, repairing the failed samples and training them, the performance issues caused by the differences in the standardization of corpus samples from different sources were resolved, and the operational capabilities of the troubleshooting intelligent agent were improved.
CN122241242APending Publication Date: 2026-06-19CHINA MOBILE INFORMATION TECHNOLOGY CO LTD +1
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA MOBILE INFORMATION TECHNOLOGY CO LTD
- Filing Date
- 2026-05-22
- Publication Date
- 2026-06-19
AI Technical Summary
Technical Problem
Existing obstacle-clearing agents suffer from poor performance during training due to differences in the normativity of corpus samples from different sources, making it difficult to effectively cope with complex and ever-changing real-world problems.
Method used
By performing format and content verification on the troubleshooting process corpus samples, the failed corpus samples are screened out and repaired to form a third corpus sample that meets the specifications, and the agent is trained based on these corpus samples.
Benefits of technology
It improves the performance of the troubleshooting agent, reduces interference and misleading during the training process, and enhances the agent's performance in operation and maintenance troubleshooting.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN122241242A_ABST
Abstract
This application discloses a method, apparatus, and electronic device for training a troubleshooting intelligent agent, applied in the field of information processing. The method includes: acquiring a first troubleshooting process corpus sample; performing format and content verification on the first troubleshooting process corpus sample to obtain a first corpus sample that passes verification and a second corpus sample that fails verification; acquiring a third corpus sample based on the second corpus sample, wherein the third corpus sample is a modified version of the second corpus sample that passes both format and content verification; and training a troubleshooting intelligent agent based on the second troubleshooting process corpus sample, which includes the first and third corpus samples. The trained troubleshooting intelligent agent is used for operation and maintenance troubleshooting. This allows training using more standardized corpus samples after verification, reducing interference from non-standard samples during training and minimizing misleading information during the training process, thereby improving the performance of the troubleshooting intelligent agent.
Need to check novelty before this filing date? Find Prior Art