A network is simply several pointsnodes that are connected to

A network is simply several points(nodes) that are connected to each other by links(edges). The term social network refers to the articulation of social relations among individuals, families, villages, communities, regions etc. Social network analysis in its abstract definition is the process of investigating social structures using network models and graph theory. Social network analysis aims to understand a community by mapping the relationships that connect them as a network and then identify key individuals and subcommunities within those communities.

Social Network analysis can even go further as to provide interpretation of the various network within large network structures.

In the 21st century, we have come to be acclaimed with two kinds of social networks namely offline social networks and online social networks. These two network categories exhibit similar properties and interpretation. Offline social networks -as the name suggests- refer to networks that exist in the real world and involve physical conduct between individuals. These include friends within a school setup, individual at a workplace etc.

Get quality help now

Dr. Karlyna PhD

Verified writer

Proficient in: Facebook

4.7 (235)

“ Amazing writer! I am really satisfied with her work. An excellent price as well. ”

+84 relevant experts are online

Hire writer

Online social networks are the relationships that exist on various internet platforms for instance friends on Facebook, Twitter as well as emails. Online Social networks mostly have a hand in glove with Social media sites (Facebook and Twitter) allowing people in different geographical locations to create and share their content over the network.

In the past decades, there have been various researches that have been carried out which have shaped our understanding of Social Network Analysis. The concept of homophily has proved to cuts across all social networks models.

Homophily suggests that people tend to have ties with people who are like themselves in socially significant ways.

In 1961 Stanley Milgram came up with the six degrees of separation theory. The theory suggests that people are connected to other individuals with a hop of at most six. If the theory still holds, this would mean that sites like Facebook with 2.27 billion (As of the third quarter of 2018) monthly active users, these users will still be having a degree of separation of less than six, probably four or five?

Across all these social networks our social ties are not all equal some are weak, and some are strong. We have strong ties with those we interact with the most, and the opposite is true. Nevertheless, this does not mean that our weak ties are insignificant.MS Granovetter-1973 proposed that there are strengths in weak ties. In the article titled Strength in weak ties, MS Granovetter argued that people we have strong ties with cannot offer us with any new information, but it is with those people we don't talk to the most we can get new information.

There has also been researching as to the spread of diseases, conventions, and news over social networks through the influence of leaders of opinions and physical conducts. According to the World Wide Web Convection 2012 from 2001 to 2011 when retweeting post users could use a various option such as 'via' or 'RT' and the 'RT' emerged to be victorious and has since become the de-facto retweeting convention.

However, most of the research mentioned above were carried out in the offline network setup, although their conclusions are also exhibited on the online networks they were still a restriction on surveys in the offline world. The introduction of Online Social Networks has opened doors to new research ideas and fields for example Data mining, Network science, natural language processing etc. These fields work with the network structures as well as the content (user data) itself. Fields like computer network and distributed systems deal with the analytics of the network structure whereas in fields like complex network theory and natural language processing there is an overlap between structure and content.

The main advantages of network analysis of on the online platform are the fact that it presents us with a mixed bag. On platforms like Twitter we can do research on billions of users with only objective restrictions to the sample space this means that we can easily access celebrities, politicians, spammers cyber-bullies etc. Online network analysis huge amounts of data are readily available. The 3V's of big data come to play here (Volume, Variety, Velocity). The Volume of the data is enormous there are petabytes of user-generated contents on Twitter alone. This data also comes in various forms such as text, sound, images etc. The velocity of the data varies from time to time and this can be predetermined for during the run-up to the USA 2016 elections more than 1 billion tweets with the #Election2016 were posted within one month. All this data can be collected automatically to be just queried when the need arises.

In the 1990s Robin Dunbar proposed what has come to be known as the Dunbar's number. Dunbar's number is a suggested cognitive limit to the number of people with whom one can maintain stable social relationships and Robin Dunbar put the number at 150.wheather this is still true on our social media platform is open for debate.

In social networks there are different kinds of graphs, these include directed graphs, undirected graphs, bipartite graphs, tripartite-hyper graphs etc. Facebook is an undirected network since if someone accepts your friend request you become mutual friends and you can both see what one will post. Twitter, on the other hand, is a directed graph since one can decide not to follow back. Sites like Flickr have tripartite-hyper graphs where every edge has three vertices including folksonomies(tags). On all these online social networks it has been established that the idea of homophily still resonates in a pattern, for example, properties like age are more homophily than gender.

The study of human behaviour according to graph properties is quite fascinating. The properties of online social network change with time. Network density varies nonmonotonically [Kumar KDD 2006]. It was also proposed that the assortative on the networks vary non-monotonically [HLL, Physics Letter A,2006]. Online Social networks are highly dynamic, with more than one new links being created every hour, but this does alter the statistical properties of the network.

Online social networks decay over time although each case is unique to the subject, for example, the decay of Myspace which was the largest social networking platform from 2005 to 2009 has not been the same for Facebook.

A different node within a network have different properties, for example, some have a high degree of centrality whereas some have a high degree of betweenness. All these nodes serve different purposes and present different interpretations. For instance, when starting a movement, you will need someone with a high degree of centrality since these people will likely be influential leaders of opinion. When measuring degree-based centrality if someone has many connections they will be important. Nodes that can easily reach others have high closeness and node that link two subcommunities have high betweenness. For example as shown in Figure.1 node A and node B have high betweenness and are called bridge nodes and certainly the shortest path will pass through node A and node B. Personally I think that nodes with high betweenness is more important than those with high degree based centrality since they link two communities and advertisers can only target these nodes and their content will reach two communities thus reducing costs.

Figure 1

Page Rank is an algorithm used by search engines to measure the authority of a webpage. Page Rank is used to rank pages according to their importance. It is generally believed that the number and importance of inbound links to a page is a significant factor. The idea behind page rank is the same as that of the importance of having links to important people in a community. Having links with important people in a community is more vital than having ties to unimportant people for example in school when one requires a recommendation letter he/she must get the letter from the Supervisor or Principal (important nodes) and not from a class monitor (unimportant nodes). Page rank will also check to see how many links a webpage has to the query.

Social network analysis also goes further as to analyse the problems that come with various networks. Nowadays, the spread of propaganda through social media platforms has become rampant and it is only through spam detection that these vices can be mitigated. In the 2016 USA election politician on the campaign trail could buy Sybil accounts to gain followers on Twitter and Facebook.

Link farming has also become a problem on our social networks. Link farming is done to fool page ranking algorithms by creating a lot of link to irrelevant pages being controlled by spammers, for example was presented by Microsoft research team suggested that in a research of about 40000 suspended accounts on twitter the average degree for random users was 36 and 234 for actual spam accounts which shows that it's just spammer following other spammers. The online social network has also presented the problem of fake news and opinions manipulation through targeted ads. With all the data in their possession platform like Facebook are been sued on cases of data breaches from their involvement with Cambridge Analytica in the 2016 USA Elections.

When events occur, there are event leaders that emerge and might be able to propagate misleading information. Social media has also changed how events are detected in the world. In natural disasters such as the 2018 Sulawesi Tsunami, it can be shown that the hashtags of the tsunami were already trending on Twitter before other people who were in the affected region had not been warned. The Qatar computer Research Institute is one of the research institutes that is currently carrying out researches on the problems of online social networks with a focus on News and Social Media Analytics, use of big data to understand urban dynamics in fast-growing cities, Extracting Timely and Credible Information from Social Media for Crisis Response etc.

In conclusion, as we continue to use online social network the line between social network and social media continues to get blurred. However, more still needs to be done to make our social networks integrated and (online networks) ethical. With all is said and done a social network be it online or offline tend to gravitate toward things like similarities of race, ideology, gender, opinions, money etc or as Smith-Lovin put it "Bird of the same feathers flock together".