Youtube Search query: “Manga”, Iterations: 5, Crawl Depth: 1
library(devtools)## Loading required package: usethis## Warning: package 'usethis' was built under R version 4.2.1devtools::install_github("mikaelpoul/readgdf")## Skipping install of 'readgdf' from a github remote, the SHA1 (1656cd91) has not changed since last install.
## Use `force = TRUE` to force installationlibrary(readgdf)nx <- read_gdf("mangachannels.gdf",as_igraph = TRUE)
nx## IGRAPH 4be971e DNW- 2423 21008 --
## + attr: name (v/c), weight (e/n), directed (e/l)
## + edges from 4be971e (vertex names):
## [1] UC-_ENWswWKMIR77czeK7Vww->UCgMU0VhhJEXaFEsFKroWI8Q
## [2] UC-_ENWswWKMIR77czeK7Vww->UC-_ENWswWKMIR77czeK7Vww
## [3] UC-_ENWswWKMIR77czeK7Vww->UCT2tPgzsKgHXpvuHGXFacnA
## [4] UC-_ENWswWKMIR77czeK7Vww->UCDgK6UKKrwcECeF-PcAd12A
## [5] UC-_ENWswWKMIR77czeK7Vww->UCESCrvsa2H5hNMmBbgkAWdQ
## [6] UC-_ENWswWKMIR77czeK7Vww->UCgmPnx-EEeOrZSg5Tiw7ZRQ
## [7] UC-_ENWswWKMIR77czeK7Vww->UCP4nMSTdwU1KqYWu3UH5DHQ
## [8] UC-_ENWswWKMIR77czeK7Vww->UCxsk7hqE_CwZWGEJEkGanbA
## + ... omitted several edgesThis network contains 2423 nodes and 21008 edges, meaning that there 2423 nodes in the network and 21008 edges illustrating how the nodes are connected. The script has created a net with 6002 videos from 250 seeds.
library(igraph)## Warning: package 'igraph' was built under R version 4.2.2##
## Attaching package: 'igraph'## The following objects are masked from 'package:stats':
##
## decompose, spectrum## The following object is masked from 'package:base':
##
## unionvcount(nx) #this shows the number of nodes/vertices## [1] 2423ecount(nx) #this shows the number of edges ## [1] 21008V(nx)$indegree <- degree(nx,mode = "in")
V(nx)$outdegree <- degree(nx,mode = "out")
V(nx)$bt <- betweenness(nx,directed=T, weights=NA)
nodelist1 <- vertex_attr(nx)
nodelist1 <- as.data.frame(nodelist1)
head(nodelist1)## name indegree outdegree bt
## 1 UC-_ENWswWKMIR77czeK7Vww 34 113 50934.75
## 2 UCCWaB2s6vQ38a8pePbhf7jQ 52 118 127046.74
## 3 UCVYe5S5_9VEQsrnLcuS1u9w 36 89 40935.88
## 4 UCe_GFJisfPSXIN8ZvHMgDJQ 28 73 80554.96
## 5 UCwGsyfyUH2S3uiTV5vES-dw 27 93 21513.63
## 6 UCQZ95oNL2aveuEp813btdyQ 45 108 465744.09A high in-degree means a high number of incoming links (or nodes that link to it). The top three nodes by in degree centrality are all music channels in Turkey. Manga is a rock band in Turkey which would explain why music videos has high in-degree centrality. I speculate that in Turkey, music channels are very popular in Turkey and a high number of channels link to it. Research showed netd müzik was the 4th most popular channel in Youtube recently and currently has the most subscribers in Turkey. The three highest in-degree nodes are:
1. UCqvzFW0qco9B_9rirRoRldg, indegree 193 connections - mor ve ötesi- a music channel in Turkey with 211K subscribers
2, UCR5wZcXtOUka8jTA57flzMg, indegree 165 connections - netd müzik- a music channel in Turkey with 23.2M subscribers
3. UCAgfTmZ3aI0iLQ6RbPKv7tw, indegree 148 connections -LeylaTheBand -a music channel in Turkey with 304K subcribersnode_in <- nodelist1[order(-nodelist1$indegree),]
head(node_in)## name indegree outdegree bt
## 289 UCqvzFW0qco9B_9rirRoRldg 193 39 46926.286
## 36 UCR5wZcXtOUka8jTA57flzMg 165 81 194849.525
## 394 UCAgfTmZ3aI0iLQ6RbPKv7tw 148 17 3066.225
## 393 UCazM5oX41f0eDAcb4N63l-w 123 41 37198.782
## 49 UCzZ1I1j--g9sFVB-6-1yzWw 113 54 82358.237
## 51 UCjT3Hl9FdnyeP7WYpbIMykA 95 100 402937.572A high out-degree means a high number of out going links (or links to other nodes). The top three nodes by in degree centrality are NOT music channels in Turkey but anime channels in the US. Manga is very popular comic and anime art in the US. I speculate that the anime community is very active in the US and has a lot of outreach in the network. Research shows that manga sales are booming in the US. The top 3 nodes by out-degree are:
1. UCG0bIjs1DXYKvaCO4t6irAQ, outdegree 163 connections - Kanon's RomCom Mangas -an anime channel in the US with 248K subscribers
2. UCtuYhirD9QNF_d0Bj-GAB9g, outdegree 125 connections - Manga Room -an entertainment channel in the US with 447K subscribers
3. UCQNbYIoWZZIgVjAkEAuAi6Q, outdegree 120 connections - Manga Sarubedo - an anime channel in the US with 92.1K subscribersnode_out <- nodelist1[order(-nodelist1$outdegree),]
head(node_out)## name indegree outdegree bt
## 10 UCG0bIjs1DXYKvaCO4t6irAQ 74 163 196915.1
## 13 UCtuYhirD9QNF_d0Bj-GAB9g 64 125 122485.7
## 22 UCQNbYIoWZZIgVjAkEAuAi6Q 56 120 215272.4
## 2 UCCWaB2s6vQ38a8pePbhf7jQ 52 118 127046.7
## 45 UCeTa378HmTWmpCFlKKIwWIQ 29 117 113418.6
## 12 UCc1lzCMSBHKA8BTjioVHpeA 42 114 204940.2The clusters seem to be about games, music videos and anime. Visually, the top nodes by in-degree in the clusters are:
1. Purple color has the largest cluster. It is about music videos and games. Visually, nodes UCrsol3PTCgS8gNcgY9C8X6w (user: Oğuz Kerem Kepenekci) and
UC6UODUn3MnYoRHBhsmgB-kg (user: MEDUSA) have high degree of centrality within this cluster
2. Blue color cluster is most connected in the community. It is an anime cluster with node UCqnskJUBid9Gip6yMutsSua (user: RomCom Manga channel)
visually showing high degree of centrality
3.
Red color cluster is small but has significant connection. This cluster is about games with node UC1-XsQ1mbEQa1Q9tNVq299w (user: Pato Horneado) having
high degree of centrality
library(visNetwork)## Warning: package 'visNetwork' was built under R version 4.2.2library(scales)## Warning: package 'scales' was built under R version 4.2.1V(nx)$size <- degree(nx,mode = "in")
wc <- cluster_walktrap(nx,weights = NULL,
steps = 4,
merges = TRUE,
modularity = TRUE,
membership = TRUE
)
V(nx)$color <- membership(wc) # set color by subgroup id
visIgraph(nx,idToLabel = TRUE,layout = "layout_nicely") %>%
visOptions(highlightNearest = TRUE, nodesIdSelection = TRUE)