Month: October 2019

The news article I will be discussing is headlined “Facebook Tests Hiding ‘Likes’ on Social Media Posts”. To summarize, Facebook began an experiment in Australia on September 26, 2019 where Likes, video view counts, and other similar metrics found on posts became hidden to users other than the original poster with the goal of reducing the significance of such measurements. In order to see if they are succeeding in improving users’ overall experience using their application, Facebook will be studying whether users would continue to comment and Like posts even when the numbers are not visible to them. An example of this can be seen in the picture provided below. Considering that the experiment is still ongoing, I will be considering three possible results that came to mind and how they relate to the material taught in the course CSCC46: Social and Information Networks. The cases that will be discussed are whether people will leave Likes and comments more often or less often, while we will not be going too far into detail about the trivial case, where there may simply be no significant changes to users’ behaviour.

As was mentioned in lectures several times, social media applications including Facebook have many different graphs in their backend that represent friendships, pages Liked, etc. These graphs are used to perform analysis to improve the application’s services such as friend recommendations and targeted advertisements.

Thus, as a result of the experiment, it is possible that users start to leave Likes and comments on posts that they otherwise would not have. For example, if a post already has thousands of comments on it, a user may decide not to leave a comment since they may think that their comment would not contribute anything to the discussion if they are too late to the post. However, if users do not know this quantity, then they may start leaving comments on posts whenever they feel like it without hesitation. This would likely lead to more connections being formed not only in real life, but in Facebook’s backend as well. Although there may be no significant immediate effect, if Facebook decides to keep the numbers hidden in the future, then there will definitely be a difference in their graphs than if they did not conduct the experiment at all. For example, although more connections may be found, a good chunk of these would likely be weak ties and some may even be negative edges, but communities would be larger in general. This change in their graphs may lead to the outputs of their machine learning and graph-related algorithms to be changed as well. For example, users may start to get more (possibly inaccurate) friend recommendations as well as advertisements for a wider variety of products.

Conversely, the experiment may lead to users to start limiting the amount of Likes and comments they leave on posts. For example, a user may start only wanting to leave Likes and comments on friends’ posts only. In this case, less connections would be formed, causing the communities found in the graphs to be more tightly-knit with more strong ties and positive edges. This may result in the previously mentioned algorithms to seem smarter since they would be more accurate in tracking our behaviours and interests. Also, let us not forget that it is possible that Facebook is selling our data, which could potentially lead to a difference in users’ experiences while using other applications as well.

In conclusion, although Facebook’s little experiment may seem trivial or insignificant, with further analysis, it has the potential to change users’ behaviour on their application. This could lead to a drastic change in the backend data as well as the services and content that people are exposed to not just on Facebook, but the other companies that they work with as well. However, it is possible that there could just be no real change in user behaviour; since this experiment is still in progress, we will see how this experiment goes but will not know the outcome until sometime in the future. Most of the ideas that were discussed here are hypothetical and are just my opinion, I happily welcome your thoughts and feedback. I also encourage you to read more into the articles referenced below if you are interested since all that was discussed here was a high-level summary and it also mentions how Instagram, another social networking service owned by Facebook, performed a similar experiment earlier this year with similar goals in mind.

Reference List:

Conger, K. (2019, September 26). Facebook Tests Hiding ‘Likes’ on Social Media Post – The New York Times. Retrieved September 27, 2019, from https://www.nytimes.com/2019/09/26/technology/facebook-hidden-likes.html

Hutchinson, A. (2019, September 27). Facebook Begins Hiding Total Like Counts on Facebook Posts in Australia | Social Media Today. Retrieved October 1, 2019, from https://www.socialmediatoday.com/news/facebook-begins-hiding-total-like-counts-on-facebook-posts-in-australia/563829/

Friendship takes up a significant portion of our social network, a good friendship can enrich our lives as well as increase your chances of happiness as an adult. However, we are not taught much about how to form a friendship. Indeed, building a relationship with someone is difficult, there are lots of factors need to be taken account of, such as education level, distance, age etc. Nevertheless, there is at least one essential requirement: You have to share something in common with that person.

Have you ever noticed how Facebook’s “People You May Know” just got creepier? In fact, If two people in a social network have a mutual friend, there lies a reasonable probability that these two people might become friends at some point in the future, this is called the triadic closure.

The dashed line indicates that B and C might potentially build a relationship. What’s more interesting behind our social network is that if we randomly choose two people across the world, we only need six or fewer people (nodes) to construct a social connection between them. On April 24, 2019, a blog on exploring your mind called “The Six Degrees of Separation Theory” by Angélica interested me. This idea was originally set out by Frigyes Karinthy in 1929 and popularized in an eponymous 1990 play written by John Guare. It is sometimes generalized to the average social distance being logarithmic in the size of the population.

The theory claimed that due to the ever-increasing network connectedness caused by technology, our world is “shrinking”. That is, even if the physical distance among two arbitrary individuals is great, the rapidly growing density of human networks made the actual social distance far smaller. If we consider a person as a node in the social network, and the number of friends they have as edges, it is not hard to imagine if a node has an extremely high node degree, then it is connected to many other nodes on the graph with a path.

In the context of CSCC46, we know that the probability that a random pair of nodes are connected on the graph is determined by the clustering coefficient, which is the quotient of the number of edges between the neighbours of a node and the degree of that node. And the average clustering coefficient is the sum of all nodes’ clustering coefficient divide by the total number of nodes. We can reveal that Karinthy’s hypothesis is true in today’s digital era. Every time when we like a post on Facebook, a video on YouTube or an image on Instagram, we somehow established a digital connection on the network, and we perform these actions constantly without realizing it. When there is more and more such digital connection on our social network, the clustering coefficient for each node increases and resulting in a higher chance that two random nodes are connected.

A social network is not only a matter of friendship, but it can also be applied to many other fields such as Theoretical Biology. José Carlos Santos & Sérgio Matos did an infodemiology study that evaluates the use of Twitter messages and searches engine query logs to estimate and predict the incidence rate of influenza-like illness in Portugal. And Vasileios Lampos proposed a method to track flu in the population using social networks. He first identified a set selected keywords to be looked for in Twitter posts, and collect a set of daily twitters, if any keyword is found in the tweet, it will be marked as 1 and 0 otherwise, and a set of equations turning statistical information into flu-score for different regions and areas.

There is still many myths about the social networks that we have not revealed its veil yet. For example, the friend paradox: people tend to have fewer friends than their friends, and homophily: the tendency for people to have (non-negative) ties with people who are similar to themselves in socially significant ways. In today’s era, it is not surprising to believe that the idea of six degrees of separation is true, and it has been several decades since this topic was popularized, who knows what number of degrees it will be now, maybe it is the same or smaller.

Reference:

Alessa, Ali, and Miad Faezipour. “A Review of Influenza Detection and Prediction through Social Networking Sites.” SpringerLink, BioMed Central, 1 Feb. 2018, https://link.springer.com/article/10.1186/s12976-017-0074-5.

Angélica. “The Six Degrees of Separation Theory and How It Works.” Exploring Your Mind, Exploring Your Mind, 23 Apr. 2019, https://exploringyourmind.com/the-six-degrees-of-separation-theory/.

Carlos, José, and Sérgio Matos1. “Analysing Twitter and Web Queries for Flu Trend Prediction.” Theoretical Biology and Medical Modelling, BioMed Central, 7 May 2014, https://tbiomed.biomedcentral.com/articles/10.1186/1742-4682-11-S1-S6.

Frei, Lukas. “Predicting Friendship.” Medium, Towards Data Science, 11 Feb. 2019, https://towardsdatascience.com/predicting-friendship-a82bc7bbdf11.

“Homophily.” Software, http://www.analytictech.com/mgt780/topics/homophily.htm.

Morse, Gardiner. “The Science Behind Six Degrees.” Harvard Business Review, 1 Aug. 2014, https://hbr.org/2003/02/the-science-behind-six-degrees.

Silver, Curtis. “How Facebook’s ‘People You May Know’ Section Just Got Creepier.” Forbes, Forbes Magazine, 11 Oct. 2017, https://www.forbes.com/sites/curtissilver/2016/06/28/how-facebooks-people-you-may-know-section-just-got-creepier/#47b743245f5a.

“Six Degrees of Separation.” Wikipedia, Wikimedia Foundation, 25 Sept. 2019, https://en.wikipedia.org/wiki/Six_degrees_of_separation.

“Tracking the Flu Pandemic by Monitoring the Social Web.” Tracking the Flu Pandemic by Monitoring the Social Web, Institute of Electrical and Electronics Engineers, 2019, https://ieeexplore.ieee.org/abstract/document/5604088.