Native vs Non-Native Language Prompting: A Comparative Analysis

TL;DR

This study compares native and non-native language prompts across multiple NLP tasks and datasets, revealing that non-native prompts generally outperform native prompts in eliciting language model capabilities.

Contribution

It provides a comprehensive analysis of prompting strategies in low-resource languages, highlighting the effectiveness of non-native prompts over native ones.

Findings

Non-native prompts outperform native prompts on average.

Mixed prompts perform better than native prompts.

Study covers 11 NLP tasks and 12 Arabic datasets.

Abstract

Large language models (LLMs) have shown remarkable abilities in different fields, including standard Natural Language Processing (NLP) tasks. To elicit knowledge from LLMs, prompts play a key role, consisting of natural language instructions. Most open and closed source LLMs are trained on available labeled and unlabeled resources--digital content such as text, images, audio, and videos. Hence, these models have better knowledge for high-resourced languages but struggle with low-resourced languages. Since prompts play a crucial role in understanding their capabilities, the language used for prompts remains an important research question. Although there has been significant research in this area, it is still limited, and less has been explored for medium to low-resourced languages. In this study, we investigate different prompting strategies (native vs. non-native) on 11 different NLP tasks associated with 12 different Arabic datasets (9.7K data points). In total, we conducted 197 experiments involving 3 LLMs, 12 datasets, and 3 prompting strategies. Our findings suggest that, on average, the non-native prompt performs the best, followed by mixed and native prompts.