文章基本信息

标题：Unsupervised Detection of Violent Content in Arabic Social Media
本地全文：下载
作者：Kareem E Abdelfatah ; Gabriel Terejanu ; Ayman A Alhelbawy 等
期刊名称：Computer Science & Information Technology
电子版ISSN：2231-5403
出版年度：2017
卷号：7
期号：4
页码：01-07
DOI：10.5121/csit.2017.70401
出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：A monitoring system is proposed to detect violent content in Arabic social media. This is a newand challenging task due to the presence of various Arabic dialects in the social media and thenon-violent context where violent words might be used. We proposed to use a probabilistic nonlineardimensionality reduction technique called sparse Gaussian process latent variable model(SGPLVM) followed by k-means to separate violent from non-violent content. This frameworkdoes not require any labelled corpora for training. We show that violent and non-violent Arabictweets are not separable using k-means in the original high dimensional space, however betterresults are achieved by clustering in low dimensional latent space of SGPLVM.
关键词：Violence; Social Media; Arabic; SGPLVM; Dimensionality Reduction; Unsupervised learning