What is Social network analysis?
Social network means to take a scenario you want to study, view what kind of interactions are happening in the system and see how to measure the interaction to answer the question we are interested in.
Social networking is not limited to just facebook friends and people and how they communicate with each other rather than it could be applied on any scenario where there are interactions. For example, we can see how under a crisis different organizations come into existence and work together towards a common goal and what kind of interaction each organization is doing among themselves. In that scenario the level of communication, volume of communication and how much information sharing is being done is important. If we see data learning we might be interested in how the instructors communicate and share information and what kind of networks are built within faculties or staff for the university. This way we can first take a question we want to answer, collect relevant information about all the entities and interactions involved and build a social network which helps us analyze the question in need.
That being said building a social network is not the only thing you need to worry about you should try to explore ways how to use techniques already build around analyzing graphs and social networks and algorithms which work well and use them to facilitate answering the question you want. What kind of measures are required for the analysis is important. You can’t just apply any measure hoping something will help you need to at least select what kind of measures can help. You don’t need a in-depth description of the measure you want to use to start with. As you use the measure more and more you will eventually would like to know more but for starters just the concept of what the measure is trying to weigh is important. Once you get more involved with social analysis you can get in-depth in a measure and learn the mathematics of that measure to be able to use it more effectively and use in new ways and more importantly interpret the results from the measure more better.
I found that taking a dataset and using Gephi was very easy. All that is required is to convert the data in a certain format which is easy to do but once the data is in form of nodes and edges you can explore the data in Gephi in even more depth and use a lot of measures and tools that can help you find things you might have not been able to find with simple algorithm analysis.
That being said not every problem can be analyzed by using Graphs and network analysis and there are limitations of some questions through the techniques i discussed. It may be that some other structure or form of analysis is needed for another type of data but don’t be afraid to explore because once you start exploring you will start understanding how to evaluate when and how to use the tools.
