Argument Mining in Data Scarce Settings: Cross-lingual Transfer and Few-shot Techniques

TL;DR

This paper investigates cross-lingual transfer and few-shot techniques for Argument Mining, revealing that data transfer and fine-tuning outperform previous assumptions about model-transfer and prompting in this complex sequence labelling task.

Contribution

It provides empirical evidence that challenges prior beliefs, showing data transfer and fine-tuning are more effective than model-transfer and prompting for Argument Mining.

Findings

Data transfer outperforms model-transfer in Argument Mining.

Fine-tuning surpasses few-shot prompting methods.

Domain and task complexity influence transfer effectiveness.

Abstract

Recent research on sequence labelling has been exploring different strategies to mitigate the lack of manually annotated data for the large majority of the world languages. Among others, the most successful approaches have been based on (i) the cross-lingual transfer capabilities of multilingual pre-trained language models (model-transfer), (ii) data translation and label projection (data-transfer) and (iii), prompt-based learning by reusing the mask objective to exploit the few-shot capabilities of pre-trained language models (few-shot). Previous work seems to conclude that model-transfer outperforms data-transfer methods and that few-shot techniques based on prompting are superior to updating the model's weights via fine-tuning. In this paper, we empirically demonstrate that, for Argument Mining, a sequence labelling task which requires the detection of long and complex discourse structures, previous insights on cross-lingual transfer or few-shot learning do not apply. Contrary to previous work, we show that for Argument Mining data transfer obtains better results than model-transfer and that fine-tuning outperforms few-shot methods. Regarding the former, the domain of the dataset used for data-transfer seems to be a deciding factor, while, for few-shot, the type of task (length and complexity of the sequence spans) and sampling method prove to be crucial.