VAST Challenge 2018 MC3- Business Data Visualizations

Team Members:

1.       Ce Pang || Student Researcher || pangc7@students.rowan.edu

2.       Dr. Bo Sun || Faculty/Supervisor || sunb@rowan.edu || Primary Contact

Analytical Tools:

1.       Tableau 2018.1

2.       Java

3. R

Estimated Work Hours:

Estimated Total Hours: 140 Hours

Public Accessibility: Granted

Video: https://youtu.be/0OP0kQuyiTU

Questions:

1. Using the four large Kasios International data sets, combine the different sources to create a single picture of the company. Characterize changes in the company over time. According to the company’s communications and purchase habits, is the company growing? Limit your responses to 5 images and 500 words.

The company is growing according to the graph above. The graph presents the accumulated (total) number of employee who participated the communications (meetings, calls, emails and purchases) over the past 2 .5 years. We separated total time period to five time frames as below and color coded in the graph:

(0). From May 11 of 2015 to December 31 of 2015

(1). From January 1 of 2016 to June 30 of 2016

(2). From July 1 of 2016 to December 31 of 2016

(3). From January 1 of 2017 to June 30 of 2017

(4). From July 1 of 2017 to December 31 of 2017

As showed, according our time frame, the employee number dropped from 13,333 to 113,518 initially, then started to increase from 113,518 all the way to 142,328. There is a record of new staff in each of the last four-time period, therefore the company is growing. The total number of staff/employee in each period is presented by “type” in the above graph.

2. Combine the four data sources for group that the insider has identified as being suspicious and locate the group in the larger dataset. Determine if anyone else appears to be closely associated with this group. Highlight which employees are making suspicious purchases, according to the insider’s data. Limit your responses to 8 images and 500 words.

To determine if anyone else appears to be closely associated with this group, we need to find someone from non-suspicious group. Thus, we filtered all the normal employees which have the communication with suspicious employees and keep the record of normal employees who have the communication more than one times.

Using 20 suspicious employees in the suspicious file, we created the first graph below to highlight normal employees (blue dot) who have communicated with the suspicious employees more than once. The x axe indicates 20 suspicious employees; y axe indicates the same time frame used in question one and the size of dots presents the number of communications.

The top 3 suspicious employees (857138, 969089 and 1690582) who had the most communications out of the group (based on dot size) are selected and analyzed using the graph below. We can see larger dots presented from non-suspicious group (normal) has indicated following employees (82 persons) who may associated with the suspicious group as well:

1376868	2040565	713892	713701
1981017	2040565	2040565	944354
175354	713429	2040565	175354
786361	713474	2040565	713474
786361	713892	713444	713489
2040565	944354	713814	713489
175354	786361	2040565	713639
713444	981554	2040565	713701
2040565	786361	175354	713701
713336	944354	2040565	713701
713410	2040565	2040565	713876
713336	2040565	713892	981554
713444	981554	713444	1963161
713876	1376868	714105	1981017
2040565	713814	2040565	2040565
2040565	714105	944354	2040565
713639	944354	981554	2040565
735108	713701	713743	2040565
713410	175354	981554	2040565
981554	713639	981554	2040565
	786361		713336

Specifically, we highlighted the employee who participated the suspicious purchase however is not in group of the suspicious employees in the graph below and whose id is 18774246.

3. Using the combined group of suspected bad actors you created in question 2, show the interactions within the group over time. Limit your responses to 10 images and 1000 words

1. Characterize the group’s organizational structure and show a full picture of communications within the group.
2. Does the group composition change during the course of their activities?
3. How do the group’s interactions change over time?

We used R to generate network graphs to present the group structures and changes. In these graphs, we labeled the edge to indicate the type of communication; each suspicious employee has a particular colorful node with id inside, the arrow means which employee initiated the communication, the width of edge means the number of communication between two employees.

The graph below shows the full picture of communication from suspicious group. It seems the group is well connected through each individual. Several persons (857138, 969089, 696047, 1690582 and 1108217) are served as center/key persons to deliver the messages. These could be the administrative staffs who are decision makers.

Over the period, the group dynamic has changed a lots as showed below. It seems that the group is very active as a whole through inter-exchange communications in 2015 and the 1^st half of 2016, however, the group became less active by involving less members and breaking down into two independent units (without communications) in the 2^nd half of 2016, then continue losing members and evolving into to 3 smaller independent units (without communications) in the 1^st half of 2017. This may due to the fact that the company is under the investigation of pollutions. Nevertheless, the group started to be active again in the 2^nd half of 2017 by having more members joined and established a larger group through inter-exchange communications.