• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Pacific-Rim conference on multimedia >Identifying Gambling and Porn Websites with Image Recognition
【24h】

Identifying Gambling and Porn Websites with Image Recognition

机译：通过图像识别识别赌博和色情网站

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Gambling and porn websites are more and more harmful to the health and growth of the youth with the rapid development of the Internet, however, the text contents and URLs based website classification methods could not get satisfying on gambling and porn websites detection because domain names of them change fast. Meanwhile, the visual based website classification has gotten perfect results in phish-ing website detection which encourages us. Therefore, we introduce the visual feature to identify gambling websites and porn websites in this paper. Firstly, we develop a website screenshot tool which could save the full contents of a website to be a image, Secondly, the effective feature is chosen by BoW model to recognize the screenshots of gambling websites and porn websites, and the appropriate parameters are chosen to promote the efficiency of classification. Finally, experimental results on our collected gambling websites and porn website datasets demonstrate that our proposed method is able to recognize the gambling and porn websites and gets satisfying results.
机译：随着互联网的迅猛发展，赌博和色情网站越来越不利于青少年的健康和成长，但是，基于文本内容和URL的网站分类方法在赌博和色情网站的检测上不能令人满意，因为他们变化很快。同时，基于视觉的网站分类在网络钓鱼网站检测中获得了完美的结果，这鼓舞了我们。因此，本文引入视觉特征来识别赌博网站和色情网站。首先，我们开发了一种网站截图工具，可以将网站的全部内容保存为图像;其次，由BoW模型选择有效功能来识别赌博网站和色情网站的截图，并选择适当的参数来提升分类效率。最后，在我们收集的赌博网站和色情网站数据集上的实验结果表明，我们提出的方法能够识别赌博网站和色情网站并获得令人满意的结果。

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