ISSS608 2016-17 T1 Assign3 Lim Hui Ting Jaclyn Conclusion

From Visual Analytics and Applications
Jump to navigation Jump to search

Intro

Approach

Findings

Conclusion

Conclusion

The usage of network graphs helps users to visualise the network in a more observable manner. With the network, one will be able to visualise the nodes that are more important, given the context. In this assignment, users will be able to see which are the key ids in the network. The use of Gephi, a platform used to visualise network graphs, allows one to not only gain an overview of the total communication data over 3 days, but also to look at how the communication data may change over time. This can be done with the use of a dynamic timeline function on Gephi. In addition, Gephi also tabulates graph metrics for the users, allowing one to gain better insight into patterns that one can observe in the network graph itself. I was able to use metrics such as betweenness centrality, eigenvector centrality, closeness centrality, on top of the degree metric.

The main results that I have observed are as follows:

  1. IDs 1278894 and 839746 stand out for their large volume of communication. ID 1278894 most likely represent a robot, that interacts with the users, i.e. through a game that requires their response at a specific time. ID 839746 is likely to be a helpline in the park.
  2. There were many communication patterns observed. As a bulk of the nodes were present in the central of the graph, this likely means that a bulk of the users were able to interact with each other. There were also other outliers, representing people who may not have had been actively communicating with the main segment of the network. These outliers are made up of people who solely interact in their own clusters, or people who only interact with external connections.
  3. The crime was likely to have been observed between 12.00pm to 12.10pm on Sunday. This was when the volume of communication was the greatest in the park.

Personally, I felt that this assignment was an interesting challenge, as it required the learning of a new software, and understanding what contributes a network graph. It is a useful takeaway in this module, as communication is commonly present in today's society. In various scenarios such as crime, one can find the perpetrator more easily by understanding communication patterns.

References

Past work by other students:

Other references: