2. Motivation
• Online social networks are expanding rapidly
• An enormous amount of research effort targets the
analysis of social networks.
• Visualizing social networks is an important aspect of
social networks research.
• We propose a visualization project which attempts to fill
some missing functionalities in pre-existing tools as
well as augment or improve others.
• This would be done from an individuals (Twitter
user’s perspective)
3. Preview of data
• Twitter is our data source -
• Friend list
• Followers list
4. Initial Analysis
• Given that the social network is a graph,
• How should this graph be traversed (breadth-first,
depth-first)
• What are the consequence of each traversal
scheme?
• Are the standard traversal schemes sufficient?
• What are the graph theoretic properties of
individual community subgraphs
6. Development Tools
• Python for algorithm prototyping/testing
• D3 for Visualization
• Php for server-side Twitter API communication
• App. only authentication for design
• 3-legged authentication for final release
8. Milestones
• Client - implemented force directed node link vis idiom
• Client - implemented adjacency matrix vis idiom
• Server - implemented 3-legged authentication
• Server - implemented breadth-first node traversal
algorithm
• Server - identified upperbound count of request to be
sent to Twitter as a variant of counting the number of
nodes in a perfect k-ary tree
• Given the request to retrieve k friends at a degree
h, the number of requests to be sent to Twitter is:
9. Milestones
• Server - identified upperbound count of request to be
sent to Twitter as a variant of counting the number of
nodes in a perfect k-ary tree
• Given the request to retrieve k friends at a degree
h, the number of requests to be sent to Twitter is: