Method for generating training course based on customer service dialogue and related device

By identifying and calculating the popularity and relevance of trending events in customer service interaction data, training courses that match actual needs are generated, solving the problem of inaccurate identification of trending events in existing technologies and improving the emergency response and business handling capabilities of customer service personnel.

CN122152970APending Publication Date: 2026-06-05CHINA MOBILE ONLINE SERVICES CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
CHINA MOBILE ONLINE SERVICES CO LTD
Filing Date
2026-01-12
Publication Date
2026-06-05

AI Technical Summary

Technical Problem

Existing technologies struggle to accurately identify real trending events from customer service interaction data, resulting in training courses that are out of touch with actual needs and fail to effectively improve customer service personnel's emergency response and business handling capabilities.

Method used

By acquiring customer service interaction dialogue data, identifying hot event sets, calculating the popularity and novelty of each hot event, selecting target hot events, and generating training courses based on these events, the course is ensured to match actual needs by using preset time decay factors and business feedback data.

Benefits of technology

It enables the accurate identification of real hot topics from conversational and fragmented customer service interaction data, and generates training courses that are precisely matched with actual training needs, thereby improving customer service personnel's emergency response and business handling capabilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122152970A_ABST
    Figure CN122152970A_ABST
Patent Text Reader

Abstract

Embodiments of the present application disclose a training course generation method based on customer service dialog and related equipment to solve the problem that the existing technology is difficult to accurately identify real hot events in the customer service scene, so that the generated training course is out of touch with the actual training needs, and cannot effectively support the improvement of customer service personnel's ability. The method comprises the following steps: obtaining a customer service interaction dialog data set, and identifying a hot event set contained in the customer service interaction dialog data set; calculating the heat and freshness of each hot event in the hot event set; selecting a target hot event from the hot event set according to the heat and freshness of each hot event; generating a training course according to the target hot event; the heat is determined based on the number of dialogues of the customer service interaction dialog associated with the hot event and a preset time decay factor; the freshness is determined based on the heat growth rate of the hot event and the ratio of the current heat to the historical average heat.
Need to check novelty before this filing date? Find Prior Art