新疆师范大学学报（自然科学版）2024，Vol.43Issue(3)：11-20,10.
全局情境约束和局部多因素融合的对话情感识别
Global Contextual Constraints and Local Multi-factor Integration for Conversational Emotion Recognition
摘要
Abstract
Conversational Emotion Recognition is a vital task in the field of Natural Language Processing(NLP),aiming to automatically identify the sentiment of each utterance in a conversation.However,most existing studies have focused on the semantic representation of utterances or modeling of speakers'relations,ignored the in-fluencing factors of emotional dynamics evolution during conversational interaction.This paper presents a new ap-proach to Conversational Emotion Recognition that incorporates global contextual constraints and local multi-factor integration.This model not only considers the global semantic information of utterances in the conversation,but also deeply explores and models multiple influencing factors of emotional semantics during the dialogue.Experimental results on public datasets demonstrate the superiority of this new method for Conversational Emotion Recognition.关键词
对话情感识别/全局情境/局部多因素融合Key words
Dialogue sentiment recognition/Global context/Local multi-factor fusion分类
信息技术与安全科学引用本文复制引用
曹卫,赵新元,薛煜阳..全局情境约束和局部多因素融合的对话情感识别[J].新疆师范大学学报（自然科学版）,2024,43(3):11-20,10.基金项目
新疆维吾尔自治区自然科学基金项目(2022D01A228) （2022D01A228）
四川省区域创新合作项目(2023YFQ0066). （2023YFQ0066）
