提升对话系统情感共情能力，实现多语言情感认可的精准识别与生成
I Understand How You Feel: Enhancing Deeper Emotional Support Through Multilingual Emotional Validation in Dialogue System
- 提出融合跨模态注意力与门控融合的多语言情感感知模型MEGUMI
- 在12万条中英日语料上实现验证响应识别与时机检测的最优性能
- 揭示大模型虽能生成多样回应，但深层情感理解仍不足
情感认可——明确承认用户感受的合理性——具有显著疗愈价值，但在计算领域研究极少。对话系统中的情感认可可分解为三部分：（i）验证响应识别，（ii）验证时机检测，（iii）验证响应生成。为此，我们发布M-EDESConv，一个包含12万条英语-日语双语对话的混合标注语料库，以及M-TESC，一个多语言语音对话测试集。针对时机检测，我们提出MEGUMI模型，通过跨模态注意力和门控融合，将冻结的XLM-RoBERTa语义与语言特异性情感编码器结合。MEGUMI在M-EDESConv和M-TESC数据集上均表现出色，兼具客观与主观优势。最后，基于EmoValidBench对GPT-4.1 Nano和Llama-3.1 8B的评估显示，当前大模型能生成上下文相似且多样的验证回应，但情感理解仍存在明显短板。
原文摘要 · Abstract (English)
Emotional validation - explicitly acknowledging that a user's feelings make sense - has proven therapeutic value but has received little computational attention. Emotional validation in dialogue systems can be decomposed into (i) validating response identification, (ii) validation timing detection, and (iii) validating response generation. To support research on all three subtasks, we release M-EDESConv, a 120k English-Japanese multilingual corpus created through hybrid manual and automatic annotation, and M-TESC, a multilingual spoken-dialogue test set. For timing detection, we propose MEGUMI, a Multilingual Emotion-aware Gated Unit for Mutual Integration, that fuses frozen XLM-RoBERTa semantics with language-specific emotion encoders via cross-modal attention and gated fusion. MEGUMI shows superior performance on both the M-EDESConv and M-TESC datasets, both objectively and subjectively. Finally, our EmoValidBench benchmarks of GPT-4.1 Nano and Llama-3.1 8B indicate that current LLMs generate contextually similar and diverse validating responses, but emotional understanding remains a major area for improvement. Project page: https://github.com/zihaurpang/Multilingual-Emotional-Validation
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。