arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2504.02178cs.CL
收藏

Subasa - 适配语言模型用于僧伽罗语低资源冒犯性语言检测

Subasa - Adapting Language Models for Low-resourced Offensive Language Detection in Sinhala

University of Kelaniya(凯拉尼亚大学);Rochester Institute of Technology(罗切斯特理工学院);George Mason University(乔治梅森大学)

Shanilka Haturusinghe, Tharindu Cyril Weerasooriya, Marcos Zampieri, Christopher M. Homan, S. R. Liyanage

AI总结:

本文针对僧伽罗语低资源冒犯性语言检测,探索了新微调策略并提出四个模型,其中Subasa-XLM-R在SOLD数据集上Macro F1达0.84,超越GPT-4o等先进模型。

摘要:

准确检测冒犯性语言对于社交媒体安全相关的诸多应用至关重要。在此任务上,低资源语言与高资源语言的性能存在显著差距。本文中,我们适配了此前未曾探索过的微调策略,将其应用于僧伽罗语冒犯性语言检测的下游任务。基于此方法,我们引入了四个模型:“Subasa-XLM-R”,它包含一个使用掩码依据预测的中间预微调步骤;“Subasa-Llama”和“Subasa-Mistral”的两个变体,分别是 Llama (3.2) 和 Mistral (v0.3) 的微调版本,并采用了特定任务策略。我们在僧伽罗语冒犯性语言检测的 SOLD 基准数据集上评估了这些模型。我们所有的模型均优于现有基线。Subasa-XLM-R 取得了最高的 Macro F1 分数(0.84),在零样本设置下于相同的 SOLD 基准数据集上进行评估时,超越了如 GPT-4o 等最先进的大型语言模型。模型和代码已公开发布。

英文摘要:

Accurate detection of offensive language is essential for a number of applications related to social media safety. There is a sharp contrast in performance in this task between low and high-resource languages. In this paper, we adapt fine-tuning strategies that have not been previously explored for Sinhala in the downstream task of offensive language detection. Using this approach, we introduce four models: "Subasa-XLM-R", which incorporates an intermediate Pre-Finetuning step using Masked Rationale Prediction. Two variants of "Subasa-Llama" and "Subasa-Mistral", are fine-tuned versions of Llama (3.2) and Mistral (v0.3), respectively, with a task-specific strategy. We evaluate our models on the SOLD benchmark dataset for Sinhala offensive language detection. All our models outperform existing baselines. Subasa-XLM-R achieves the highest Macro F1 score (0.84) surpassing state-of-the-art large language models like GPT-4o when evaluated on the same SOLD benchmark dataset under zero-shot settings. The models and code are publicly available.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑