Introducing Blanche

1
Introducing Blanche
Science of Networks in Communities (SONIC) Team
Engineering Collaboratory University of Illinois
at Urbana-Champaign April, 2005
2
Elements of Blanche Nodes

• Nodes individuals, groups, organizations,
  databases
• Attributes of nodes attitude, age, gender,
  organizational affiliation, expertise,
  human/non-human agents
• - Attributes can be continuous, categorical,
  constant or changing

3
Elements of Blanche Relations

• One or more relations between the nodes
• communication
• information flow
• access, publish or retrieve information
• perceptions of others attitude or expertise
• Relations can be directional, non-directional,
  binary, continuous, constant or changing

4
As Attributes Attitude, Expertise
F
A
E
As Relations Communication from A to F
D
B
C
5
Explaining changes Attributes of the nodes

• Changes in the value of the attributes for actor
  i are influenced by present and prior values of
• Relations of actor i with other actors j
• Attributes of other actors j
• Other attributes of actor i

6
Explaining changesRelations between the nodes

• Changes in the value of relations for actor i to
  actor j are influenced by present and prior
  values of
• Attributes of actor i
• Attributes of other actors j
• One or more relations between actor i and all
  other actors j

7
Example

• To model the relationship between peoples
  communication with one another and the similarity
  in their attitudes towards two issues X and Y

8
Generative Mechanisms

• Actor is attitudes towards X (or Y) is based on
  the attitude of other actors, j, with whom actor
  i communicates.
• Actor is communication with j is based on their
  previous communication and shared attitudes
  towards X and Y.

9
Variables

• Attribute Variables
• AttX Attitude towards X. Range 0 to 1
• AttY Attitude towards Y. Range 0 to 1
• Relational Variables
• PComm Probability of a communication tie
• Range 0 to 1
• Comm Presence of a communication tie
• Range 0 or 1

10
Process Preview

• Create and specify variables
• Enter equations
• Check levels
• Specify file names for initial data sets
• Generate and save new data sets

11
Process Preview

• Model analysis
• Run simulation single or multiple runs
• View results Attribute and Relational data, Time
  series graphs, Visualizations
• Analyze results Descriptive univariate and
  network statistics
• Save and export results text files, UCINET DL or
  Krackplot KP files.

12
Viewing Nodes and Variables
Click here to enter a new variable
Click here to view and modify highlighted variable
13
Viewing and Modifying Variable AttX Attitude
about Topic X
Click here to generate new initial datasets
14
Generating Data for Initial Attitudes Towards
Topic X AttX
15
Previewing Data for Initial Attitudes Towards
Topic X AttX
Click here to save the data generated
16
Saving New Data for Initial Attitudes Towards
Topic X AttX
Click here to save the data generated
17
Viewing and Modifying Variable AttY Attitude
about Topic Y
18
Viewing and Modifying VariableProbability of a
Communication tie PCOMM
19
Viewing and Modifying Variable Presence of a
Communication tie COMM
20
Running model in BlancheStep 1 Specify model
parameters
21
Running model in BlancheStep 2 Click on Play
button
Click here to run simulation
22
Running model in BlancheStep 3 Allow simulation
to run
23
Running model in BlancheStep 4 View data for
selected attributes
AttX and AttY at iteration 99
24
Running model in BlancheStep 5 View data for
selected relations
Click here to go to previous iteration
Comm at iteration 97
25
Visualizing data in Blanche Select to graph
individual actors or the entire network by
highlighting the values tab and selecting the
add button
26
Visualizing Graphs Graph of the data (normal
view)
Zoom area
27
Visualizing Graphs Graph of the Data (3D view)
28
Visualizing Networks at iteration 30
29
Saving data
30
Possible applications

• Examine change in simulation results as a
  function of
• Variance in the initial conditions
• Different equations based on different theories
  or different interpretations of the same theory
• Use empirical data as initial data to predict
  future states of the network attitudes and
  communication relations.
• Use results to guide research design when and
  what data is most appropriate to test hypotheses