用量化大模型与翻译增强跨语言情感检测，表现优异。
UWB at WASSA-2024 Shared Task 2: Cross-lingual Emotion Detection
- 量化模型+低秩适配器微调，结合多语言Transformer
- 机器翻译提升双任务性能，数值触发词检测第一
- 适合跨语言情感分析、多语种应用研究者参考
本文介绍为WASSA-2024跨语言情感检测共享任务构建的系统。任务包含两个子任务：一是从六类情绪中为五种语言的推文打标签，二是以二元和数值形式预测引发情绪的关键词。我们提出的方法基于量化大语言模型（如Orca~2）与低秩适配器（LoRA）微调，结合多语言Transformer模型（如XLM-R、mT5）。通过机器翻译增强两个子任务性能，并在第二子任务中使用触发词切换策略。系统在数值触发词检测中排名第一，在二元触发词检测中排名第三，在情绪分类中排名第七。
原文摘要 · Abstract (English)
This paper presents our system built for the WASSA-2024 Cross-lingual Emotion Detection Shared Task. The task consists of two subtasks: first, to assess an emotion label from six possible classes for a given tweet in one of five languages, and second, to predict words triggering the detected emotions in binary and numerical formats. Our proposed approach revolves around fine-tuning quantized large language models, specifically Orca~2, with low-rank adapters (LoRA) and multilingual Transformer-based models, such as XLM-R and mT5. We enhance performance through machine translation for both subtasks and trigger word switching for the second subtask. The system achieves excellent performance, ranking 1st in numerical trigger words detection, 3rd in binary trigger words detection, and 7th in emotion detection.
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。