After a course that basically flew by, I look back on all of the things that we covered, and I see an instant connection to the organization where I work.I work at a medium-sized university in a rural state. We have been struggling with things like enrollment, retention, graduation rate and of course finances. Through all of the different committees and sub-committees, much is said about what we ought to be doing to make and impact on these different issues and objectives. The interesting thing is that nobody is talking about using the data that we have or could obtain access to as a source helping make intelligent, data-driven decisions. Even more interesting is the fact that we as a university are trumpeting that introduction of a new "Big-Data" program. However, we don't seem to be incorporating any of the ideas or principle about Big-Data, Business Intelligence and what they could do to help this university prosper again.
During the most recent module where we were looking at simple graph theory and network analysis and visualization, it was pretty obvious how much can be ascertained with only a little effort and some decent tools. Obviously, Gephi was a great network analysis and visualization tool; however, NetVizz for Facebook did a lot of the heavy lifting. Being able to create the necessary data-structures for Gephi was hugely important for it to be able to work its magic on the data.
When one gets the opportunity to step back and look at the entire stack of what we covered in 587, you can readily see how all of the tools and practices that we touched were dependent on the first few modules. For example, Gephi was hugely dependent on our ability to create data it could read from Facebook or LinkedIn. Microstrategy provided by Teradata was hugely dependent on having a data-warhouse behind the tool. Google Analytics also manages to store incredible amounts of important data in a way that makes BI tools easy and useful.
The Ted talk given by Hans Rosling put into perspective how powerful data visualization really is. The slides on the top ten data visualizations were also eye opening. Then, beginning to apply some of these concepts to our own data using Gephi and to some degree Google Analytics, showed how visualization successfully can perform one of the most important tasks in a data-driven world. It connects people with their data and communicates the trends, interactions, relation in ways that people can digest and understand.
Network visualization using Gephi personally showed me more about the people in my Facebook than I actually knew about them. In addition, it showed me a few things about myself that I knew, but Gephi showed how they impacted me over time. For example, I don't participate all that frequently on Facebook with my friends/acquaintances; thus, I don't really pay that close attention to what they "Like" and therefore what their life preferences are. However, looking at my acquaintances' likes, it was easy to see which ones were liberal, which ones are conservative, which ones are musicians, etc. To be honest, it was much quicker to view them in this manner than it would ever have been to read through all of their profiles on Facebook.
Running some of the analysis on my network, I saw those who were the lynchpins in my network, meaning those that were the nodes that if eliminated segregated my network into multiple sub-networks. I saw who among them was the most connected within his sphere (and part of my sphere); interestingly, it wasn't someone that I would have expected.
This information made me think about how graph and network theory could be applied in the organization where I work. Much of what we need to understand about the market we are in (higher education) involves the relations about our students, where they came from, what they do when they are in school, where they go and so on. Looking at this type of data visually, I'm sure that we could more easily wee where our students are coming from, where they are not coming from, what things they are interested in, etc. to establish better marketing efforts, and perhaps a better understanding of what types of things attract them into staying through graduation.
When I back up and look at all of the material we covered in this course and try to connect it to the university where I work, I see the number one need for our organization is building up our overall understanding of the data we have available and working to construct a data warehouse capable of supporting the institutions data needs. It became clear throughout the course that the real power is not merely in the tools but in the proper understanding of what data one has and how to best structure it so that the tools can be used effectively to analyze it. Hopefully, our university will begin to follow its own advice and begin to implement a strategy for understanding its data by applying the principles we learned for Big-Data and business intelligence.
