Day 12 of #100daysofnetworks
Announcing book giveaway winners!
On Day 8 of #100daysofnetworks, I announced a book giveaway. In order to participate, someone just needed to come up with a topic of interest, use the wikipedia crawler to create a knowledge graph, and then do whatever network analysis they felt comfortable doing, based on their current level of experience.
When I decided to write my book “Network Science with Python”, my goal was to inspire even just a few people to dive into the topic, knowing that it would not receive anywhere near as much attention as Machine Learning. That’s really unfortunate, because I build a lot with ML, and I find this even more useful. Even knowing that this would be a quieter topic, I needed to write this book, to show how these techniques can be useful.
So, I am very excited to announce today’s winners of the current book giveaway, because they did something very special. They took a leap into the unknown, to learn more, to expand their understanding, to expand the boundaries of the possible. I respect that!
For the competition, I stated that I would give away five copies of my book. If there were more than five submissions, I’d do a raffle. There were three submissions, so the raffle is not needed! Every single entry is a winner. Let’ get to it and congratulate the winners!
Winner #1 - Josh Leigh - Causal Knowledge Graph
Josh Leigh used the Wikipedia Crawler to create a knowledge graph related to Causal Inference and Discovery. With Aleksander Molak’s new book on Causal Inference and Discovery having been recently released, there is qiute a lot of excitement on the topic. There is also overlap with Graphs and Causal Inference. We will be exploring Causal Discovery throughout #100daysofnetworks, for sure.
Josh Leigh even made a cool LinkedIn post, showing some of his work! Check it out!
There’s a couple quotes I especially liked:
“A quick look into Endogeneity and my mind is blown - we see this in mineral system geoscience and it's rarely addressed. It's one reason why I’m interested in learning more about causality.”
I like this, because he made a discovery and a connection that related to his actual work. Networks are great learning tools, and he will be able to use this to dig deeper.
“Straight away I see reference to System Dynamics which is linking through to Causal loop diagrams! I have an understanding of these so this is super helpful to build from those concepts. I can also see a whole bunch of things that sound super interesting but I have no clue about! What the heck is Confirmatory factor analysis!? 🤯”
That’s the good stuff. Every time I look at a network, I’ll come out with some new insights. One network visualization will be jam packed with insights, and it’s impossible to find them all in one run. Now he can explore confirmatory factor analysis, and then that may lead him to new places.
I loved also loved Josh’s visualization of the core of the network.
Excellent work, Josh Leigh! Congratulations!
Winner #2 - Iftikhar Ud Din - Game of Thrones Social Network Analysis
Iftikhar Ud Din made a great submission to the book giveaway in the form of Social Network Analysis of the Game of Thrones Social Network. I have been curious about this network for a while, but never found time to find the dataset. Iftikhar pointed me to it, so I will add it to the “Datasets” page on substack. We’ll use that dataset in #100daysofnetworks, as it looks really, really fun. Learning should be really, really fun.
For now, the dataset can be found here!
Iftikhar did a really cool analysis, using things in networkx I had never seen before. He has inspired me to learn more!
I really like that he showed the degree distribution. I haven’t been doing that, but have wanted to bring it up. I will do a post on that at some point.
The reason that’s important is because networks tend to follow a Power Law, where few nodes will have many edges and many nodes will have few edges. Take a look. Many nodes have less than ten edges (degrees, connections). It looks like only one node has more than sixty edges. This is common in real-world networks, and it is interesting to see in fictitious social networks, as well. Perhaps it has to do with an author’s ability to manage characters. You can learn more about this phenomenon here.
Iftikhar did an excellent job and went above and beyond, even doing Egocentric Network Analysis and some community detection. I look forward to seeing where these capabilities take him! Great job! Congratulations!
Winner #3 - Yuval Feinstein - Georgia Knowledge Graph
Yuval Feinstein made a Knowledge Graph related to the country of Georgia. He built this Knowledge Graph using the Wikipedia Crawler and used the country name of Georgia as well as several cities. He used Page Rank to identify important nodes (Wikipedia Pages), and these four were found to be most important:
Georgia
Politics of Georgia
Economy of Georgia
Religions of Georgia
That makes a lot of sense, intuitively. Religion, politics, and the economy have importance to culture, identity, and prosperity.
He found one issue: due to the ambiguity of the name Georgia, Wikipedia pages relating to the state Georgia were included in the network. This is definitely something to keep in mind, and it’s similar to a google search. Sometimes, you need to include additional words to filter things OUT, or if this graph were to be used for something, then preprocessing could be done to remove the Georgia (State) related nodes.
Everyone, Great Job, and Thank You!
I had five books ready for the giveaway, and I’ll give away three! I’ll keep the rest around for future giveaways. There will be more! We’re only on day eleven! Thank you, Josh Leigh, Iftikhar Ud Din, and Yuval Feinstein for participating! You are the winners! Congratulations. I will be in touch with how to get access to your free copy of my book!
If you would like to learn more about networks and network analysis, please buy a copy of my book!