Network analysis
We would talk about the influential flow graph.
This analysis includes data from AllMusic.com.
To dig into the influence with various musicians. We put each influencer and follower as the nodes and the influential flow as the edges from the influence data. Finally we get a directed graph with 5568 nodes and 42761 edges.
The graph attribute is the following:
.jpg)
The modularity is between from 0.3 to 0.7 so this graph is a significant influential flow graph. Since there are a bunch of edges, the density and average betweeness are small. The average degrees is over 10, which means a musician still intersects with many others.
We pick out the top ten musicians with highest betweeness and degrees, which are shown below.
We pick out three famous musicians based on degrees and draw their influential graphs. We think there has been main influential leap in the music world. You can see that Bob Dylan influences many musicians like The Beatles and The Rolling Stones and these artists continue to influence others. Their influence spread in different genres.
Eventually, We draw the two scatterplots for degrees, betweeness and cluster coefficient to explore their correlations.
 |  |  |
|---|
It shows that degree and betweeness is not correlated but degree and cluster coefficient are negatively correlated. While if someone is more influential, he would have a smaller cluster coefficient. That’s because his influence spreads in various genres and those musicians would influence less with each other. Otherwise, the musicians within a smaller particular genre would have higher possibility to influence others more.