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 2005.11424cs.CL
收藏

基于transformer的社交媒体会话线索中的上下文感知讽刺检测

Transformer-based Context-aware Sarcasm Detection in Conversation Threads from Social Media

Emory University(埃默里大学)

Xiangjue Dong, Changmao Li, Jinho D. Choi

摘要:

我们提出了一个基于transformer的讽刺检测模型,该模型考虑了整个会话线程的上下文,以实现更强大的预测。我们的模型使用深层的Transformer层来执行线程中的目标话语和相关上下文之间的多头注意力。在来自社交媒体Twitter和Reddit的两个数据集上对上下文感知模型进行了评估,并显示出比基线提高了3.1%和7.0%。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,成为该共享任务中36名参与者中性能最高的系统之一。

英文摘要:

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.

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。

↑