ImplicitAVE
收藏资源简介:
ImplicitAVE是首个公开的多模态数据集,专门用于隐式属性值提取,由伊利诺伊大学芝加哥分校创建。该数据集包含68,000个训练实例和1,600个测试实例,覆盖五个不同领域,旨在解决现有数据集在处理隐式属性值方面的不足。数据集通过精心策划,包括了产品图像和文本信息,以支持多模态学习。ImplicitAVE的应用领域包括电子商务中的产品表示、推荐和分类,特别是在需要从产品图像和文本中推断隐式属性值的场景中。
ImplicitAVE is the first publicly available multimodal dataset dedicated to implicit attribute value extraction, created by the University of Illinois Chicago. It contains 68,000 training instances and 1,600 test instances covering five distinct domains, and is designed to address the shortcomings of existing datasets in handling implicit attribute values. The dataset has been carefully curated to include both product images and textual information to support multimodal learning. Application scenarios of ImplicitAVE include product representation, recommendation and classification in e-commerce, particularly in cases where implicit attribute values need to be inferred from product images and text.




