AssignYL Analysis

From Visual Analytics and Applications
Jump to navigation Jump to search

YL title bg.jpg VAST 2018 Mini Case 3: Net of Justice Has Long Arms


BACKGROUND

DATA PREPARATION

EXPLORATIONS

NETWORK ANALYSIS

CONCLUSIONS

Narrow the Suspects

Findings Visualization Evidence
Closeness & Betweeness


With pieces of information gained from the exploration part, we may tell that there are people who are involved in suspicious connections without any interactions. So for the first step of the analysis, we may filter out people by degree less than 6.

  • Closeness: Closeness represents the average shortest path scores from one node to another. The higher closeness of a node, the more center a node is. People with higher closeness centrality in the graph tend to have faster influence over the network. As we can tell from the graph of node size plotted by closeness, Meryl Pastuch, Lindsy Henion, Richard Fox, and Jose Ringwald are people with relatively high closeness, and Melita Scarpaci is someone with high closeness despite the low weight.


  • Betweenness: Betweenness measures the number of times a node has lied on the shortest path between 2 other nodes. The higher betweenness of a node, the more influential to the flow system that node is. This is implying that Lindsy Henion, Jose Ringwald, Richard Fox and Tobi Gatlin - the group of people with high betweenness - may be the ones who hold authority within a group.


Speculations:
Melita Scarpaci - a influencer but maybe from higher management
Richard Fox - a practioner, a popular staff in Kasios, a leader for his group
Meryl Pastuch, Lindsy Henion - leaders of their groups, but their team may have closly work with Richard's team
Rosalia Larroque & Kerstin Belveal - staffs work for the purchasing group, they are involved with all of the suspicious purchases and their email communications are the strongest among all


Analysis Period: 2015-2017

YL NW 2 Closeness.PNG

YL NW 2 Betweeness.PNG


Tracking by Years


YL NW 3 2015.PNG


Who's behind this?


Filter by nodes degress greater than 6 will simplify the graph.