Analyzing Meme Uniqueness
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4175728/#s1
In the modern digital age, a phenomenon resulting in the development of ‘cultural units defined by atomic concepts’ (as stated by Michele Coscia) has radically impacted the content and spread of information and culture in mainstream media today. This, of course, refers to the creation and distribution of what are known as memes, and can be modeled with regards to social networks and information cascades. Memes strive to win the attention of people as a heterogeneous social network in competition for this limited resource. The article cited above formulates a method and analysis on how uniqueness contributes the popularity of a meme and its likelihood to go viral.
The above network is comprised of nodes that represent memes, with their size proportional to the average number of ‘votes’, or likes, received per time-step. Orange nodes have more votes per time-step, while those in blue received less votes. The network visualization is of a matrix S, calculated using a method that assigned a numerical value associated with the uniqueness of a meme. This network is comprised of a single giant component, and shares qualities with other real world complex networks such as average edge density.
The results of this network visualization are very easy to see. For example, we can notice that more popular memes are located at the perimeter of the network, while nodes more centralized are less successful/popular. Edges are formed between memes if they share similarities, and we can see that they have the same color scheme and positioning as the nodes. Furthermore, this network is created without information regarding the social networks of creators/uploaders, revealing an meme independence not necessarily rooted in high power members of a social network, but rather on meme content and originality. One last observation that can be made is the remarkable similarity to biological systems and other common network systems that this meme network has. Similar to biological systems recording the passing down of genes, the meme network displays how original ideas generate new ones, leading to the creation of new and popular memes.
Lastly, the meme network, can be modeled as an information cascade given the four basic properties that they are comprised of.
- The decision to make for a user is what kind of meme they should post in order to receive the most ‘votes’ and win the most attention.
- The user makes decisions in a sequence based off his own history of meme posting.
- The information that helps a person decide what kind of meme to post comes from that available and ‘trending’ on social media. The number of likes these posts have received as well as the comments are information that will help the user evaluate his decision.
- Lastly, the trends of memes on social media alongside access to archives and history allows the user to see what earlier decision makers did, but not necessarily information about themselves and what they knew.
Through all of these analyses, we see how memes are a very insightful modeling subject for social networks and information cascades.