• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Business information systems workshops >Feature Extraction Based on Semantic Sentiment Analysis
【24h】

Feature Extraction Based on Semantic Sentiment Analysis

机译：基于语义情感分析的特征提取

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The abundance of information available on the Web raises the need for techniques that are able to analyze and make a better use of such huge information. Extracting features from unstructured text and assign for each feature its associated sentiment in a clear and efficient way is the goal of this paper. Aspect or feature sentiment analysis is the suitable level of sentiment classification especially for dealing with the domain of products and their related features. We introduce two important and text-related techniques for this purpose namely sentiment classification and semantic Web technology. Sentiment analysis extracts people's opinions in an automatic manner, while semantic Web technology provides a useful improvement by enriching data with additional meaning (semantically annotated) and makes product features better understood by human and automatically processed by machines. For evaluation, we apply our sentiment analysis method twice. Once on a plain corpus, and once again on the same corpus after applying semantic annotation to its data. We also compute the precision and recall principles to evaluate our information extraction method over the corpus in its both cases.
机译：Web上可用的信息很多，因此需要能够分析和更好地利用这些巨大信息的技术。本文的目标是从非结构化文本中提取特征，并为每个特征分配清晰相关的情感。方面或特征情感分析是合适的情感分类级别，尤其是在处理产品领域及其相关特征时。为此，我们介绍了两种重要的与文本相关的技术，即情感分类和语义Web技术。情感分析以一种自动的方式提取人们的意见，而语义Web技术则通过丰富具有附加含义的数据（符号注释）来提供有用的改进，并使人们更好地理解产品特征并由机器自动处理。为了进行评估，我们两次使用了情感分析方法。一次在普通语料库上，然后在对数据应用语义注释后再次在同一语料库上。我们还计算精度和召回性原则，以评估我们在两种情况下对语料库的信息提取方法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号