A user emotion recognition method based on text information, medium and device

By using text-based user intent recognition and emotion contagion models, combined with a spatial distance weighting function, the problem of inaccurate emotion recognition in digital customer service systems when voice and video input are lacking is solved, achieving highly accurate emotion assessment and service response in privacy-preserving scenarios.

CN121958554BActive Publication Date: 2026-09-11MOBILE TECH COMPANY CHINA TRAVELSKY HLDG
4 Cites 0 Cited by

Patent Information

Application Number
CN202610077627.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-01-21
Publication Date
2026-09-11
Estimated Expiration
2046-01-21

AI Technical Summary

Technical Problem

Existing digital customer service systems, lacking voice and video input, rely solely on text content for emotion analysis, resulting in low accuracy and an inability to effectively capture users' true emotional states. This is especially problematic in high-density crowds where they cannot distinguish between shared and unshared anxieties, leading to mismatched service responses and a reduced user experience.

Method used

By using text-based user intent recognition and emotion contagion models, combined with spatial distance weighting functions, the influence of individual and group emotions is calculated. A two-stage emotion assessment mechanism is designed to distinguish the contagion weights of common and non-common anxieties, thus compensating for the shortcomings of emotion recognition under a single text modality.

Benefits of technology

In privacy-protected scenarios, it improves the accuracy of emotion recognition and the quality of service response, and can accurately compensate for users' emotional assessment when multimodal input is lacking. It significantly enhances the emotional perception ability of digital customer service, especially in providing timely and effective reassurance and solutions in the event of emergencies.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The present application relates to the technical field of user emotion recognition in digital customer service, in particular to a user emotion recognition method based on text information, medium and equipment. The method comprises the following steps: determining the influence period and range of emotion contagion based on the current time and location; combining the basic emotion score and the emotion score of other users meeting the conditions to calculate the comprehensive negative emotion value. The formula considers the emotion score of other users, spatial distance weight and intention weight, and the high anxiety common intention weight is higher. The present application accurately models the emotion contagion effect through the spatial distance weight function, distinguishes the contagion weight of common and non-common anxiety, improves the accuracy of emotion evaluation under single text mode, assigns higher weight to common anxiety events, and enhances the emotion contagion effect. Its influence will be significantly amplified, thereby compensating for the underestimation of the target user's emotion caused by single text input.
Need to check novelty before this filing date? Find Prior Art