• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Computers in Human Behavior >Online reputation measurement of companies based on user-generated content in online social networks
【24h】

Online reputation measurement of companies based on user-generated content in online social networks

机译：基于在线社交网络中用户生成的内容的公司在线声誉评估

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

Social media websites such as Facebook, Twitter, etc. has changed the way peoples communicate and make decision. In this regard, various companies are willing to use these media to raise their reputation. In this paper, a reputation management system is proposed which measures the reputation of a given company by using the social media data, particularly tweets of Twitter. Taking into account the name of the company and its' related tweets, it is determined that a given tweet has either negative or positive impact on the company's reputation or product. The proposed method is based on N-gram learning approach, which consists of two steps: train step and test step. In the training step, we consider four profiles i.e. positive, negative, neutral, and irrelevant profiles for each company. Then 80% of the available tweets are used to build the companies' profiles. Each profile contains the terms that have been appeared in the tweets of each company together with the terms' frequencies. Then in the test step, which is performed on the 20% remaining tweets of the dataset, each tweet is compared with all of the built profiles, based on distance criterion to examine how the given tweet affects a company's reputation. Evaluation of the proposed method indicates that this method has a better efficiency and performance in terms of recall and precision compared to the previous methods such as Neural Network and Bayesian method. (C) 2015 Elsevier Ltd. All rights reserved.
机译：诸如Facebook，Twitter等社交媒体网站已经改变了人们交流和决策的方式。在这方面，各种公司都愿意使用这些媒体来提高声誉。在本文中，提出了一种信誉管理系统，该系统通过使用社交媒体数据（特别是Twitter的推文）来衡量给定公司的声誉。考虑到公司的名称及其相关的推文，可以确定给定的推文对公司的声誉或产品具有负面或正面影响。所提出的方法基于N-gram学习方法，包括两个步骤：训练步骤和测试步骤。在培训步骤中，我们考虑四个概况，即每个公司的正面，负面，中立和不相关的概况。然后，将80％的可用推文用于建立公司资料。每个配置文件都包含出现在每个公司的推文中的术语以及术语的频率。然后，在对数据集的剩余20％推文执行的测试步骤中，根据距离标准将每个推文与所有已构建的配置文件进行比较，以检查给定推文如何影响公司的声誉。对所提出方法的评估表明，与诸如神经网络和贝叶斯方法之类的先前方法相比，该方法在查全率和精确度方面具有更好的效率和性能。 （C）2015 Elsevier Ltd.保留所有权利。

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