arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1802.07862cs.CL
收藏

用于短社交媒体帖子的多模态命名实体识别

Multimodal Named Entity Recognition for Short Social Media Posts

Carnegie Mellon University(卡内基梅隆大学);Snap Research(Snap研究院);Intuit(Intuit公司)

Seungwhan Moon, Leonardo Neves, Vitor Carvalho

摘要:

我们引入了一个名为多模态命名实体识别(MNER)的新任务,用于嘈杂的用户生成的数据,如推文或Snapchat标题,其中包括带有图像的短文本。这些社交媒体帖子经常出现不一致或不完整的语法和词汇符号,周围的文本上下文非常有限,给NER带来了重大挑战。为此,我们为MNER创建了一个名为SnapCaptions的新数据集(Snapchat图像-标题对提交给公共和众包故事,带有完全注释的命名实体)。然后,我们建立在最先进的Bi-LSTM基于词/字符的NER模型基础上,1)一个深度图像网络,它结合了相关的视觉上下文来增强文本信息,2)一个通用的模态注意力模块,它学习衰减不相关的模态,同时放大信息量最大的模态来提取上下文,适应每个样本和令牌。建议的MNER模型与模态注意力显着优于国家的最先进的纯文本NER模型,成功地利用提供的视觉环境,打开潜在的应用程序的MNER在无数的社交媒体平台。

英文摘要:

We introduce a new task called Multimodal Named Entity Recognition (MNER) for noisy user-generated data such as tweets or Snapchat captions, which comprise short text with accompanying images. These social media posts often come in inconsistent or incomplete syntax and lexical notations with very limited surrounding textual contexts, bringing significant challenges for NER. To this end, we create a new dataset for MNER called SnapCaptions (Snapchat image-caption pairs submitted to public and crowd-sourced stories with fully annotated named entities). We then build upon the state-of-the-art Bi-LSTM word/character based NER models with 1) a deep image network which incorporates relevant visual context to augment textual information, and 2) a generic modality-attention module which learns to attenuate irrelevant modalities while amplifying the most informative ones to extract contexts from, adaptive to each sample and token. The proposed MNER model with modality attention significantly outperforms the state-of-the-art text-only NER models by successfully leveraging provided visual contexts, opening up potential applications of MNER on myriads of social media platforms.

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。

↑