Identity and Deception in the Virtual Community
Summary:
In the article, Donath studies how identity is established in online social platforms, effects of identity deception, concealment & impersonation and conditions giving rise to these phenomena by performing an ethnographic study on Usenet newsgroups. She suggests that an individual’s identity cues are spread all over Usenet’s letter in the form of email address (especially the domain name), user’s signature (name, age, sexual orientation, title etc), language and writing style. She argues that these cues are not always reliable and can be faked by trolls. They can impersonate full identities or engage in category deception (like age and gender deception). She talks about how trolling can negatively influence the dynamics of the group by undermining the feeling of trust in a community by giving several examples. The author concludes by saying that design of the online platform plays a key role in identity establishment and deception.
Reflections:
Identity does play a very important role in social media. What the author suggested in this 1944 article still holds today. Our identity cues are spread over social media platforms including but not limited to our name, display picture, online friend network, educational information, phone number, the language we use, the images we share. Online platforms allow users to create multiple accounts each exhibiting same or different persona. So, can a fake account be linked to a real account? Can a user be linked to all his social media accounts? Does a user’s accounts, fake or real exhibit some similar features and characteristics? I believe a lot of research is already happening to answer these questions. Similarly identifying fake social media accounts is also a hot topic these days. The author suggests that design of a system can influence the identity and identity cues. Several in-domain models have been proposed to detect a fake account but is it possible to have a generic model that can work effectively on several social media platforms together? Also, what steps can social media platforms take to ensure authenticity of a user’s profile? E.g. In India, several matrimonial websites have started Aadhar verification to validate a user.
All social media platforms are fighting these problems with no full proof solution yet. A news article[1] stated that Facebook deleted 583 million fake accounts in the first quarter of this year using it’s AI tools. Google released Perspective, a tool that detects online abuse in 2017. I believe it’s very important for the social media platforms to reveal the algorithms they deploy so that researchers can study and improve them.
Anonymity has given birth to trolls and cyber bullies. How are their profiles and daily activity different than other users? Do their posts have linguistic markers that differentiate them from the rest? It would be interesting to study the research methods that exist to detect troll profiles, cyberbullies and their posts. What has been accomplished and what more needs to be done. The paper suggests use of high costs on deception. Some punishments like banning a user exist. But these don’t stop the user to create another account. What other punishments can be introduced? What will be the implications of these punishments?
Links: [1] https://www.digitaltrends.com/social-media/facebook-fake-accounts/
4chan and /b/: An Analysis of Anonymity and Ephemerality in a Large Online Community
Summary:
In the paper, authors investigate the effects of concepts: anonymity and ephemerality in online communities. They do so by performing two large scale studies [consisting of 5,576,096 posts in 482,559 threads collected over a period of two weeks] on 4chan’s most popular board, ‘random’ or /b/ board. They provide examples of 4chan’s popularity [~7 million users] and influence on Internet culture [origination of several popular memes like LOLcats] and several highly-visible, off-site activities where 4chan members participated [hacktivist group Anonymous]. The authors state that although offensive content [nudity, racism, homophobia] is the part of 4chan’s identity, the platform is also a source of several funny and creative memes. They start by analyzing the content posted on /b/ by collecting 598 posts over ten days and come up with nine high level categories that reflect the themes of the posts that start threads on /b/. ‘Themed’ and ‘Sharing Content’ were the most popular categories which revolved around sharing images and web content in keeping up with /b/’s identity as the image bulletin board. They next quantify ephemerality by collecting a large dataset of activities comprising 5 million threads. In the first study, the author finds that most threads spend 5 sec on the first page and are short lived, having a lifetime of less than 5 min. Users use two phenomena: Bumping and Saging to influence ephemerality. Next the authors study identity signals and anonymity. /b/ allows users to posts in three ways: ‘Anonymous’ly, any random name or tripcode. Majority of the posts were found to be anonymous and a negligible amount were using tripcodes to guarantee identity. Instead people post their pictures with timestamps while others use slang specific to the platform to claim high status and identity in the community.
Reflections:
The paper studies two important design phenomenon ‘anonymity’ and’ ephemerality’ in large online communities. Ephemeral content seems to be one of the emerging social media trends which is quite prevalent in the increasing popularity of snapchat video that also inspired other social media platforms to mimic this feature, e.g Whatsapp Status feature, Facebook & Instagram stories. Authors suggests that ephemeral platforms lead to fresh content every time but whether that is always a quality content needs to be thought about. Also, an interesting research would be to find out if people get addicted more to the ephemeral platforms than the ones where data is kept permanently. I feel as the content on such platforms is temporary, a user might feel a need to visit these platforms more often.
The paper suggests that some threads lasted for as long as 6 hours and some posts have 519 replies. We can study what all factors make a post stay for a longer duration on this platform and make it popular. Is it only content? Originality? or the language? Also, the dataset under observation had few instances of bumping and sagging. Extended study can be performed on larger data spreading over a larger window to actually know their impact on ephemerality.
A question that comes to my mind is if the content on social media really ephemeral? Like the paper itself mentions that there are several websites like 4chanarchive.org that archive the website’s posts and threads. Also, content of a user’s interest can always be saved locally in his computer. It would be interesting to see how many old posts/memes get reposted on the platform.
Also are we really anonymous online? Most of the times an anonymous user can be traced using the IP Address of the device or the geolocation data.
An interesting study would be to collect data posted anonymously on various similar websites and study what type of content usually gets posted anonymously. How is that tied to the culture of the online community? Is the anonymous content linguistically different from the one posted with identity?
It was interesting to see how users of 4chan adopted practices to signal identity. What motivates users to claim partial identity? Does it only happen when their posts start receiving attention? One factor that the paper mentions is to communicate high status. One can study what other visual, textual or linguistic cues are given by anonymous users intentionally or unintentionally on other similar social media platforms. Considering the ephemeral nature of posts are there any users who were able to persist their identity on the platform?
The paper gives rise to a debate: Identity or Anonymity? While anonymity can encourage intimate conversations, people can also use it to cyberbully which can cause unimaginable harm to victims. E.g. anonymity empowers users to ask personal questions on mediums like quora without the fear of being judged. While Sarahah was banned after accusations against the app breeding haters and bullies. I personally advocate neither and believe that people should always have the option to choose either of the two.
Reflections on data used in the paper:
The authors in the paper have introduced nine high level categories to depict the theme of posts on /b/. Having fine grained categories could have provided a richer view of the content. Also, the relative frequency of the appearance of these categories was done on a small window. The authors could have used the same two-week window that they used to analyze ephemerality and anonymity.
I am not sure how do the authors conclude that demographics of /b/ are mostly North Americans? Why was EST considered? Is figure three plotted only using activity of users in North America?
