ISSS608 2017-18 T3 Assign Vishal Badrinarayanan Conclusion

From Visual Analytics and Applications
Jump to navigation Jump to search
Breaking-chain1.jpg

Overview

Data Preparation

Question 1

Question 2

Question 3

Question 4

Conclusion

BACK TO HOME

TO UNCOVER INFORMATION BY BREAKING THE NETWORK CHAIN

    1. We can find that the company is growing at a steady growth.There was less activity in the initial months but then after October they was an increase in the number of employees in the company.
    2. All the other forms of communication except meeting didn't have a lot of increase.

    1. We find that Richard Fox and Meryl Pastuch are the most influential people in the network.
    2. Gail Feindt is the one who is involved in most of purchases in the network.
    3. We found other probable suspicious nodes in a network based on their communication pattern and with whom they are communicating.
    4. We find that the people belonging to dark green cluster are the least suspicious.

    1. Different form of communication types and whether it is within their group or across different groups determine their influence.
    2. There are many groups in the network and the factors like Betweenness centrality tells about how connected a node is a network , but for determining the organizational structure it is important to understand the type of communication between networks. For eg Lindsey Henion in red cluster has higher centrality but will be lower ranked than Rosalia and Kerstin.
    3. In a dynamic graph we can identify the changes in the nodes across years much better than a static graph. Here the change in size or color can help us determine changes. In this we found that there was a very high amount of activity initially and it only increased again by 2017 end.
    4. The influential nodes like Richard Fox,Meryl were highly active at the start.After that they did not have much of an increase in activity and again became actively involved by mid-2017. This is shown by change in node size based on centrality.