Blog 4 | Week 6 | Social Media Analytics and Network Analysis
Introduction
Whenever I hear about networks, I immediately think about a nature video that I saw on mycelial networks and their symbiotic relationships. It is the plant version of social networks. It's so enthralling that Star Trek has an episode of traversing through the fungal network--a concept encompassing beyond the 4IR fusion of physical, digital, and biological spheres to include navigational travel?!?! Here is the YouTube video for anyone that wants to take a quick break and understand how to get an all access travel pass through "the network."
About Module 3
Module 3 starts with an Intro to Networks, then details to Network Visualizations and Quantitative Network Metrics. I assume Community Detect and Interpretation is next. Lecture 11 begins with the understanding that networks consist of nodes (vertices) and relationships (edges), which can be single or two mode. Edges can be weighted/unweighted and/or directed/undirected. And social media is based on ego networks. Barabasi in his Thinking in Network Terms video, states that the networking phenomenon is a result of so many people having so many devices, and that the communications on those devices are stored as data that can be used to build theories because the connectedness can be quantified and described. Once you have theories, you can mathematically formulate and have predictive power on social behaviors. Wow.
Lecture 12 is about the various types of networks, i.e., the different kinds of layouts that are graphically created to help us explore, communicate, and understand data. Network visualizations help us do this and it can be seen in various forms:
 |  |  |  | |
| Forced Directed | Geographic | Circular | Clustering | Hierarchical |
Lecture 13 is about network properties - centrality measures and others i.e., density, clustering coefficient, reciprocity, and cliques. In this article, Social Network Analysis of Current NBA Players on Twitter, Ogeleka (author) examines the Twitter posts of NBA players. He uses general network properties to identify who is the most popular among colleagues, their paths of connectedness, how players out-degree, following, correlate to their in-degree, being followed. Through this, he also identifies the antisocial players--those who aren't reciprocating followers.
My understanding of this module is that until recently, there wasn't a way to analyze people's behaviors whether it's communicating with others or examining their consumption of goods and services. Network visualizations, or network graphs, are for casting light on how relationships are with lots of different entities. It helps scientists, managers, etc. understand who or what are the entities in a network and how are they connecting with each other. Then by applying some mathematical equations, the scientists or managers can apply predictive measures.
After reading the examples on our Facebook group of different network analyses, I found the Network of Thrones, Song of Math and Westeros, to be the most interesting and probably most relevant to this module (so thank you Ethan Panal for finding and posting!). I would love to see a geographic network of Westeros. The breakdown of the centrality degrees was really helpful because it verifies what we kind of already know about the characters, and how Tyrion stands out in so many ways.
In conclusion we know that networks aren't new, but the way that we can now look at stored data and present the findings graphically is relatively new. There are all kinds of networks to evaluate that can help organizations understand their consumers and their products or services better.
Many thanks to my classmates for reading my blogs and Facebook posts. Feel free to connect with me in LinkedIn. See you in the next class. -Kat Francisco
Resources:
Ogeleka, C.H. (2020). Social Network Analysis of Current NBA Players on Twitter. Available at:
https://medium.com/analytics-vidhya/social-network-analysis-of-current-nba-players-on-twitter-b3fb9a741806
https://medium.com/analytics-vidhya/social-network-analysis-of-current-nba-players-on-twitter-b3fb9a741806
Hi Kat,
ReplyDeleteGreat blog, it was very insightful and the Star Trek clip was an interesting application of network visualization. This module was definitely interesting and it is crazy to see the application of network visualization.
Thanks Dylan! I agree that the network graphs are interesting.
DeleteAGREED!!! Great blog post, and the Star Trek reference to the tardigrade having full access to the network was NOT lost on me! What a fantastic analogy because network analysis really was explained well in the clip. Great work and summary!
ReplyDeleteThanks Adrienne! I can't imagine how people ever got through school without Google.
DeleteWhat a great Star Trek clip and summary of this module! I found this module very interesting, and it has gotten me thinking about how I could've used some of the knowledge I've gained from this module in some of the analyses I've done in the past.
ReplyDeleteThanks Leon! We worked with clusters in another class, but it would have been nice to see the connectedness. I wish there was another class that we could dive more into Gephi and Tableau. I felt this way with the classes on Python, SQL, and R too.
DeleteHi Kat,
ReplyDeleteAt the risk of sounding redundant, I too was a fan of the star trek clip! This module was something that was very new for me and also really interesting. I wasn't very familiar with social media and network analysis; The numerous interesting use cases, from the lecture and from the social media posts were so exiting to read.
I loved reading your blog, great summary :)
Thanks Yashree! I hoped it wasn't too much out-of-the-box... it's all about getting from point A to point B, right? lol.
DeleteThank you for your wonderful review, Kat! It's interesting how after this module we're finding applications of networks in sports, TV shows, pop culture, etc. They're all around us! Side note: how did you make that table visual with the different types of networks, in blogger?
ReplyDeleteThanks Loren! For the visual graphs, I used the html function on blogger and put each picture in a table 95% width --> table row --> table data 20% width. :-)
DeleteHi Kat, I really liked the way you laid out your visuals within this blog! The article you linked regarding the popularity of NBA players was an interesting read. I like to think that network analyses like this can be combined with other KPIs to create interesting outputs. For example, does how likable you are influence how many passes you may get in a game? Does it affect your overall contributions to the team? Does a more likable player tend to generate more assists?
ReplyDeleteWay to circle back to KPIs! I almost forgot about that. There must be a correlations between player success on the court and on social media. I mean, I wouldn't follow a terrible player. It seems harsh but it's true.
DeleteHello Kat, thanks for your entry. Great explanation of the lectures, I agree this week was really interesting and I enjoyed learning about networks as well. I didn't realize that complex systems like networks could be understood using quantitative measures. Thanks for your blog post!
ReplyDeleteThanks Reynaldo!
Delete