论文标题
社交媒体的对话线程中基于变压器的上下文感知讽刺检测
Transformer-based Context-aware Sarcasm Detection in Conversation Threads from Social Media
论文作者
论文摘要
我们提出了一个基于变压器的讽刺检测模型，该模型从整个对话线程中说明了上下文，以进行更强大的预测。我们的模型使用深层变压器层在目标话语和线程中的相关上下文之间执行多头关注。在社交媒体，Twitter和Reddit的两个数据集上评估了上下文感知的模型，并比其基线显示了3.1％和7.0％的改进。我们的最佳模型分别为Twitter和Reddit数据集提供了79.0％和75.0％的F1分数，成为此共同任务中36位参与者中表现最高的系统之一。
We present a transformer-based sarcasm detection model that accounts for the context from the entire conversation thread for more robust predictions. Our model uses deep transformer layers to perform multi-head attentions among the target utterance and the relevant context in the thread. The context-aware models are evaluated on two datasets from social media, Twitter and Reddit, and show 3.1% and 7.0% improvements over their baselines. Our best models give the F1-scores of 79.0% and 75.0% for the Twitter and Reddit datasets respectively, becoming one of the highest performing systems among 36 participants in this shared task.
