top of page

Are your news comments violating military censorship?

Dr. Inbal Yahav presents a new study on Censorship Violation through Commenting in Social Media at the 11th Symposium on Statistical Challenges in eCommerce, Addis Ababa, in Ethiopia.

Scholars of digital democracy, journalism, new media, and social networking have overwhelmingly embraced the grand vision of improved public discourse through user participation in online news discussions. This growing body of research focuses, to a large degree, on factors that influence participation, studying factors related to the news subject matter, the network characteristics of users, the demographic characteristics of users, and more. The underlying goal motivating much of this research lies in the belief that more participation, deeper engagement, higher levels of awareness, and larger networks with stronger ties are all “good for democracy” or as Dahlberg states the “widespread enthusiasm about the possibility of digital media technology advancing and enhancing democratic communications”.

We study participation behaviors in the specific context of source-censored online news articles and the comment discourse that they arouse. Our approach to the study of comment participation in the context of censored news articles begins with trying to understand the underlying social and behavioral relationships and extends into the realm of detection. The need to detect and prevent certain forms of commenting behavior has already been recognized in the context of profanity and sexually offensive language, and profanity filters abound [2]. Technological solutions to the latter are relatively straightforward based on text analysis at the time of comment input. Detecting censorship circumvention through comments is a far more complex task involving the understanding of context, social network structure, and news article content. This complexity grows when we consider that the comment may not be intentionally breaching censorship and only provide clues which, when systematically analyzed, reveal the censored element.

bottom of page