UWB at WASSA-2024 Shared Task 2: Cross-lingual Emotion Detection

TL;DR

This paper describes a system for cross-lingual emotion detection in tweets, utilizing fine-tuned multilingual language models and translation techniques, achieving top rankings in trigger word detection and strong performance in emotion classification.

Contribution

The paper introduces a novel approach combining quantized large language models with low-rank adapters and machine translation for improved cross-lingual emotion detection.

Findings

Ranked 1st in numerical trigger word detection

Ranked 3rd in binary trigger word detection

Ranked 7th in emotion classification

Abstract

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.