论文标题
多模式仇恨言论的分类 - 仇恨模因挑战的获胜解决方案
Classification of Multimodal Hate Speech -- The Winning Solution of Hateful Memes Challenge
论文作者
论文摘要
仇恨模因是多模式分类的新挑战，重点是检测多模式模因中的仇恨言论。添加了困难的示例，以使很难依靠单峰信号，这意味着只有多模型才能成功。根据基拉（Kiela）的说法，与人类（64.73％vs. 84.7％的准确性）相比，最新方法的性能较差。我提出了一个新模型，将多模式与规则相结合，该模型的准确性和AUROC的第一个排名分别为86.8％和0.923。这些规则是从培训集中提取的，并专注于提高困难样本的分类准确性。
Hateful Memes is a new challenge set for multimodal classification, focusing on detecting hate speech in multimodal memes. Difficult examples are added to the dataset to make it hard to rely on unimodal signals, which means only multimodal models can succeed. According to Kiela,the state-of-the-art methods perform poorly compared to humans (64.73% vs. 84.7% accuracy) on Hateful Memes. I propose a new model that combined multimodal with rules, which achieve the first ranking of accuracy and AUROC of 86.8% and 0.923 respectively. These rules are extracted from training set, and focus on improving the classification accuracy of difficult samples.
