arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2408.04540cs.CL
收藏

MemeMind 在 ArAIEval 共享任务中:通过说服技巧识别定位阿拉伯语文本中的说服性片段

MemeMind at ArAIEval Shared Task: Spotting Persuasive Spans in Arabic Text with Persuasion Techniques Identification

Hamad Bin Khalifa University(哈马德·本·哈利法大学);Northwestern University in Qatar(卡塔尔西北大学)

Md Rafiul Biswas, Zubair Shah, Wajdi Zaghouani

AI总结:

本文提出基于 AraBERT-base 的两阶段微调方法,对阿拉伯语推文和新闻文本进行 BIO 序列标注以识别宣传技巧片段,在 ArAIEval 任务 1 中以 F1 0.2774 获得第 3 名。

摘要:

本文聚焦于检测来自推文和新闻段落的阿拉伯语文本中的宣传性片段和说服技巧。数据集中的每个条目包含一个文本样本和相应的标签,这些标签指示文本中宣传技巧的起始和结束位置。落在已标注片段内的词元被分配为“B”(开始)或“I”(内部)、“O”,对应于具体的宣传技巧。利用注意力掩码,我们为每个片段创建了统一长度,并根据提供的标签为每个词元分配 BIO 标签。然后,我们使用 AraBERT-base 预训练模型进行阿拉伯语文本分词和嵌入,并配合词元分类层来识别宣传技巧。我们的训练过程采用两阶段微调方法。首先,我们仅训练分类层若干轮,随后进行全模型微调,更新所有参数。该方法使模型能够适应宣传检测任务的特定特征,同时利用预训练 AraBERT 模型所捕获的知识。我们的方法取得了 0.2774 的 F1 分数,在任务 1 的排行榜上获得第 3 名。

英文摘要:

This paper focuses on detecting propagandistic spans and persuasion techniques in Arabic text from tweets and news paragraphs. Each entry in the dataset contains a text sample and corresponding labels that indicate the start and end positions of propaganda techniques within the text. Tokens falling within a labeled span were assigned "B" (Begin) or "I" (Inside), "O", corresponding to the specific propaganda technique. Using attention masks, we created uniform lengths for each span and assigned BIO tags to each token based on the provided labels. Then, we used AraBERT-base pre-trained model for Arabic text tokenization and embeddings with a token classification layer to identify propaganda techniques. Our training process involves a two-phase fine-tuning approach. First, we train only the classification layer for a few epochs, followed by full model fine-tuning, updating all parameters. This methodology allows the model to adapt to the specific characteristics of the propaganda detection task while leveraging the knowledge captured by the pre-trained AraBERT model. Our approach achieved an F1 score of 0.2774, securing the 3rd position in the leaderboard of Task 1.

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。

↑