Diversity Over Size: On the Effect of Sample and Topic Sizes for Topic-Dependent Argument Mining Datasets

TL;DR

This work investigates the effect of TDAM dataset composition in few- and zero-shot settings and shows that, while fine-tuning is mandatory to achieve acceptable model performance, using carefully composed training samples and reducing the training sample size can still yield 95% of the maximum performance.

Abstract

The task of Argument Mining, that is extracting and classifying argument components for a specific topic from large document sources, is an inherently difficult task for machine learning models and humans alike, as large Argument Mining datasets are rare and recognition of argument components requires expert knowledge. The task becomes even more difficult if it also involves stance detection of retrieved arguments. In this work, we investigate the effect of Argument Mining dataset composition in few- and zero-shot settings. Our findings show that, while fine-tuning is mandatory to achieve acceptable model performance, using carefully composed training samples and reducing the training sample size by up to almost 90% can still yield 95% of the maximum performance. This gain is consistent across three Argument Mining tasks on three different datasets. We also publish a new dataset for future benchmarking.

Figures (7)

• Figure 1: Sample experiments on the fsc
• Figure 2: Sample experiments on the iam
• Figure 3: Sample experiments on the ibm
• Figure 4: Topic experiments for FS150T-/IAM- and IBM-Corpus on ERNIE 2.0 and FLAN-T5 XL and in F${_1}$ macro.
• Figure 5: Sample experiments for FS150T-/IAM- and IBM-Corpus on ERNIE 2.0, FLAN-T5 XL, Llama2-70B, and ChatGPT in F${_1}$ macro and with standard deviation.