According to Dr. Tronick, human has a nature to make a relationship with others since they were born. We know it as social relationship. A number of toddlers who meet somewhere will somehow end up in conversation and play together even though there are no instructions from the adults. Adults who have the same hobby or interest will gather in a community.
In Bandung City, we can find a wide range of communities for al- most every hobby. For ex- ample, in Jalan Taman Sari, we can find the secretariat of Yamaha NMax scooter owner community. In Jalan Purnawarman, we can find a place where people who love and learn French cul- ture gather together.
In Pasteur, especially in a mall, we can see people play or have a regular com- petition held by the com- munity of Tamiya car fans. Of course, there are many other similar locations in Bandung.
Does a community make it compulsory to provide a physical location as a place to gather? It is apparently not.
By the end of 20th cen- tury, Indonesian students who were studying abroad formed an online commu- nity, namely KasKus (now it is famous for its trading activity). In the beginning of 2000s, the owner of Apple made a community in the form of mailing-list apple- id. There are also photo- graphy fans that established a photographer community dot net.
A community, especially online community, almost barely recognizes organiza- tional structure. Although a community has one or more people who role as pioneers, but, during the activity, it is the contributions from the members that make the community alive.
If there is no formal regulation in a community, can we measure the perfor- mance of the community? How? This is where we are introduced to graph theory.
Before that, let us go to Königsberg City (It is now called Kaliningrad). In the 18th century, there was a cha- llenge to cross 7 bridges of Königsberg. Not only to cross it, but also to finish the cha- llenge, someone must pass the bridges once.
Being curious with the solution, a mathematician, Leonard Euler showed the blueprint of bridges into a graph. Through this visualiza- tion, Euler showed that the challenge was impossible to resolve.
Euler tried to
represent
Representation of 7 Bridges of Königsberg in the form of nodes
and edges (common. wikimedia.org)
(see figure 3b). it is a simple
Method, isn’t it?
The other answer is: Yes, since digital record from every activity in social media sim- plify the task of the researcher to collect relational data of internet users. Social media becomes one of the dominant factors to make SNA method repeat its popularity.
The activity in social me- dia can be interpreted in a graph by, for example, symbol- izing the user as a node, and an edge to connect that user
the bridges and lands connect- ed by nodes (dots) and edges (lines). Land was symbolized by node, and bridge was sym- bolized by edge. So, if a land was connected to another land by a bridge, there would be a node and another node that was connected with a line (see figure 2). The solution that was invented in 1936 is be- lieved to be the basic of graph theory, one of the branches of math.
Now, how can we use the graph theory in a community? We can use the same analogy as one used by Euler. If Euler used a node and an edge to symbolize a land and a bridge, we can use a node and an edge to symbolize people (or inter- net users), and identify the correlation that might happen. In social life, this graph theory becomes the basic of a method namely Social Net- work Analysis (SNA). Is it a breakthrough in social media era? The answer is yes and no. No, because the use of SNA was already applied by Jacob Moreno (a psychiatrist) in 1930s, to analyze the cor- relation of friendship in an elementary school. It utilized a very simple way; it observed the activity of students in that elementary school during break time, and Moreno noted the activity of the students in that elementary school.
Every student was sym- bolized by a node. If a student played or talked to other stu- dents, Moreno drawed a line (an edge) from a node that represented that student to another node that represented his friend. After he drawed it into a graph, Moreno could see the most popular student. For example, by counting who had the most edges in the graph.
The other answer is: Yes, since digital record from every activity in social media sim- plify the task of the researcher to collect relational data of internet users. Social media becomes one of the dominant factors to make SNA method repeat its popularity.
The activity in social me- dia can be interpreted in a graph by, for example, symbol- izing the user as a node, and an edge to connect that user with his uploaded messages or with other users who are actually his friends in that so- cial media.
For example, figure 3 shows the representation of a graph from social inter- action that was taken from Telkom University facebook page. Here, the nodes repre- sent facebook users that par- ticipated in that page, and also messages contained in that page. While the edges repre- sent the relationship between facebook users and messages. For example, if a user writes a message in Telkom University facebook page, there will be a line (an edge) between two nodes that represent the user and the message that he wrote. Also, when someone does “like” or “share” a message.
The problem is when a network consists of more than ten thousand nodes, and they are connected, how can we analyze them? Visually, the representa- tion of the graph as seen in 3a might not talk much. But there is a piece of good news.
Graph theory provides a number of matrices that can be used to analyze complex networks. Some most-used matrices, such as central- ity (who the central actor is, or what the most impor- tant node is in a network), modularity (how a group or community is shaped in a network), and density (how dense the relation among the nodes in a network). And there are still many other SNA matrices.
Instead of doing manual calculation, just like how it was done by Jacob Moreno, algorithm software can do the task for us. It is just because the speed of the algorithm is limited by the power of hardware (in this case, it depends on the pro- cessor power). The calcula- tion speed is also related to network complexity that will be analyzed.
In social network, if the number of the nodes in a network increases, the com- plexity level of the network also increases exponentially. With a huge number of nodes, say millions, it is important to have more processors to count SNA matrix faster. One of ways to fasten the counting with limited processor is by optimizing the existed algo- rithm.
One of ways that has been done is by doing graph com-pression. We merge same nodes and edges to become supernodes and superedges. The aim is to reduce graph complexity that finally fastens the calculation process.
The result shows that the network data that is saved (for some cases) can be reduced up to 40%, and the calcula- tion process of SNA matrix is faster up to 50%. Every effort in simplifying the graph will surely lead to the decrease of the accuracy of the calculation result, and this is a work that is included in the agenda of our next research.
Efficient algorithm, in the
end, will make analysis pro- cess cheaper. The availabil- ity of abundant and real-time social media data opens op- portunity for stakeholders to make a more agile organiza- tion.
Imagine if we can monitor the responses from a commu- nity of our marketing program in real-time. Or, we can pre- dict staff’s performance in an organization without doing time-consuming survey. Orga- nization policy can be more effective, and organization can have backup resources that can be allocated for more im- portant aspects.
The key is how an organi- zation adopts this technology fast, and makes it a competi- tive advantage .
Cited from a research: “Penyederhanaan Kompleksitas Meterik Centrality dalam Large Scale Social Network di Bidang Ilmu Manajemen” by Yahya Peranginangin, S.T., MBA and Andry Alamsyah, S.T., MBA.
