UR-FUNNY
收藏资源简介:
UR-FUNNY是一个用于理解幽默的多模态语言数据集,由罗切斯特大学和CMU的研究团队创建。该数据集包含8257个幽默和非幽默的实例,涵盖文本、视觉和声学三种模态。数据来源于TED演讲视频,通过分析演讲者的语言、表情和声音来识别幽默。创建过程中,研究者利用了TED视频的转录和观众反应标记,提取了幽默的上下文和关键点。UR-FUNNY数据集的应用领域主要集中在自然语言处理中,旨在通过多模态分析解决幽默识别的问题。
UR-FUNNY is a multimodal language dataset for humor understanding, developed by research teams from the University of Rochester and CMU. This dataset contains 8257 humorous and non-humorous instances, covering three modalities: text, vision and acoustics. The data is sourced from TED talk videos, where humor is identified by analyzing the speaker’s language, facial expressions and vocal features. During the dataset construction, researchers utilized the transcripts and audience reaction tags of TED videos to extract humorous contexts and key points. The UR-FUNNY dataset is mainly applied in the field of natural language processing, aiming to solve the problem of humor recognition through multimodal analysis.

- 1UR-FUNNY: A Multimodal Language Dataset for Understanding Humor计算机科学系,罗切斯特大学,美国;语言技术研究所,SCS,CMU,美国 · 2019年



