Summary:
The article “Antisocial Behavior in Online Discussion Communities” analyzes undesirable user participation in online communities, and how to detect them early on before they are banned from communities. The goal of this is to minimize troll-like behavior, which results in more positive online communities. There are already methods of trying to prevent this, such as reporting posts, down voting, and blocking. Although these methods are still in place, there is still a large amount of trolling in the online communities. The study uses three online discussion-based communities: CNN.com, Breitbart.com, and IGN.com. Through this, users were categorized as Future-Banned Users (FBUs) or Never-Banned Users (NBUs). While going through the users’ behaviors, the article evaluates three main questions regarding how and when users start deviant behavior online. NBU’s and FBU’s can be analyzed to find out more information about whether or not someone will be banned. Design features such as post content, user activity, community response, and actions of community moderators help do this. The Never-Banned Users and Future-Banned Users are split by the rate in which their posts are deleted. Depending on how fast a post is deleted, it can be predicted whether or not the account should be taken down.
Reflection:
The actions of the Future-Banned Users imply that they will produce deviant behavior, even before referring to them as “FBU’s”. From using social media sites such as Twitter, most of the deviant content that I come across does not make much sense. The irregular tweets are usually a mix of inappropriate words replying to a previous tweet or being stated. Most Future-Banned Users write much differently than accounts that I follow, and it is easy to spot out. As discussed in the article, most deviant content is concentrated within individual threads. What if the threads are private? Is there an efficient way to monitor that without intruding someone’s privacy? I do think that this is a major issue within online social communities, but it may be harder to solve with privacy issues.
Questions:
Is there a way to prevent deviant content that is private?
Why do people feel the need to post deviant tweets?
