ATE数据集
收藏arXiv2017-09-27 更新2024-08-06 收录
下载链接：
http://arxiv.org/abs/1709.09220v1
下载链接
链接失效反馈官方服务：
资源简介：
ATE数据集是由瑞士联邦理工学院洛桑分校人工智能实验室创建，用于方面术语提取的研究。该数据集包含23296条来自亚马逊的评论数据，旨在通过远监督方法提高方面术语提取的准确性。创建过程中，研究者利用注意力机制筛选出包含实际意见方面的句子，从而提高数据集的质量。该数据集主要应用于方面情感分析，解决传统情感分析中无法识别具体目标的问题。
The ATE (Aspect Term Extraction) dataset was developed by the Artificial Intelligence Laboratory at École Polytechnique Fédérale de Lausanne (EPFL) for research on aspect term extraction. This dataset comprises 23,296 pieces of review data collected from Amazon, with the goal of improving the accuracy of aspect term extraction through distant supervision methods. During its construction, researchers employed the attention mechanism to screen sentences that contain genuine opinion aspects, thus elevating the overall quality of the dataset. This dataset is primarily utilized in aspect-based sentiment analysis, resolving the limitation that traditional sentiment analysis fails to identify specific target entities.
提供机构：
瑞士联邦理工学院洛桑分校人工智能实验室
创建时间：
2017-09-27
