2. RED/082311
Keen Analytics
Agenda
• What is a graph?
• The connection between behavior and graphs
• What can graphs tell us about behavior?
• Interactive Graph Visualizations
2
4. RED/082311
Keen Analytics
What is a Graph?
4
1
2
3
4
5
0
Vertex/Node Edge
Cycle
Self Edge
Articulation Point
Degree = 2
In Degree = 0
Out Degree = 2
Degree = 4
In Degree = 3
Out Degree = 1
Degree = 1
In Degree = 1
Out Degree = 0
Degree = 5
In Degree = 2
Out Degree = 3
5. RED/082311
Keen Analytics
Behavior and Graphs
5
As we interact with the world around us,
we experience physical, mental, and virtual connections.
Every smile, touch, click, email, joke, and IM makes connections.
Connections may last an instant, a lifetime, or multiple life times.
Collectively, these connections form a graph of all interactions.
Social media, email, blogs, etc., form part of this graph
6. RED/082311
Keen Analytics
How to make sense of graphs
How are events connected?
What are the communities?
What is unusual?
(Anomaly Detection)
Who is important/influential? Are two patterns the same?
(Isomorphism)
What can be inferred?
(Semantics/Inference)
6
20. RED/082311
Keen Analytics
Example: Detect Abnormal Behavior
Degree Centrality = 3
(Vertex edge count)
Probability
Degree Centrality
20
Rare Occurrence
(Frequent
Communication)
Rare Occurrence
(Infrequent
Communication)
Normal
Communication
Levels
Editor's Notes
This is a slide build showing the visualized results of:
Centrality analysis: To identify the most central entities in your network, a very useful capability for influencer marketing.
Path analysis: To identify all the connections between a pair of entities, useful in understanding risks and exposure.
Community detection: To identify clusters or communities, which is of great importance to understanding issues in sociology and biology.
Sub-graph isomorphism: To search for a pattern of relationships, useful for validating hypotheses and classifying subgraphs
Anomaly Detection: Detect and characterize unusual patterns of behavior to search for abnormal situations, such as hacker attacks or fraud
Behavior Prediction: Graphs represent behavior over time and contain behavior patterns. Identifying partial patterns enables behavior prediction.
Visualization of Social Graph Density.
Briefly describe what Social Density could be used for
The following slide shows how the was created
This slide portrays how the Social Graph Visual in the previous slide was created
Visualization of Centrality Distribution.
Briefly Describe what Centrality Distribution could be used for
The following slide shows how the was created