Extract Agent-based Model from Communication Network - PowerPoint PPT Presentation

About This Presentation
Title:

Extract Agent-based Model from Communication Network

Description:

Title: Extract Agent-based Model from Communication Network Last modified by: Justin Document presentation format: On-screen Show Other titles: Cochin ... – PowerPoint PPT presentation

Number of Views:136
Avg rating:3.0/5.0
Slides: 26
Provided by: rpi81
Learn more at: http://www.cs.rpi.edu
Category:

less

Transcript and Presenter's Notes

Title: Extract Agent-based Model from Communication Network


1
Extract Agent-based Model from Communication
Network
  • Hung-Ching (Justin) Chen
  • Matthew Francisco
  • Malik Magdon-Ismail
  • Mark Goldberg
  • William Wallance
  • RPI

2
Goal
Given a societys communication history, can we
  • Deduce something about nature of the society
  • e.g., Do actors generally have a propensity to
    join small groups or large groups?
  • Predict the societys future
  • e.g., How many social groups are there after 3
    months?
  • e.g., What is the distribution of group size?

3
General Approach
Individual Behavior (Micro-Laws)
Learn
Societys History
Predict (Simulate)
Societys Future
4
General Approach
Individual Behavior (Micro-Laws)
Learn
Predict (Simulate)
Societys Future
5
Social Networks
  • Individuals
  • (Actors)
  • Groups

6
Social Networks
  • Individuals
  • (Actors)

1
2
- Join
- Leave
  • Groups

3
7
Social Networks
  • Individuals
  • (Actors)

1
- Join
- Leave
  • Groups

- Disappear
- Appear
- Re-appear
3
8
Societys History
9
General Approach
Individual Behavior (Micro-Laws)
Learn
Societys History
Predict (Simulate)
Societys Future
10
Modeling of Dynamics
11
Example of Micro-Law
Actor X likes to join groups.
SMALL
LARGE
Parameter
12
ViSAGEVirtual Simulation and Analysis of Group
Evolution
State Properties of Actors and Groups
Decide Actors Action
State
Normative Action
State
State update
Actor Choice
State
Process Actors Action
Feedback to Actors
Real Action
13
General Approach
Individual Behavior (Micro-Laws)
Learn
Societys History
Predict (Simulate)
Societys Future
14
Learning
?
Learn
?
15
Groups Group Evolution
16
Actors Types
  • Leader prefer small group size and is most
    ambitious
  • Socialite prefer medium group size and is medium
    ambitious
  • Follower prefer large group size and is least
    ambitious

17
Learning Actors Type
  • Maximum log-likelihood learning algorithm
  • Cluster algorithm
  • EM algorithm

18
Testing Simulation Data
19
Testing Real Data
20
General Approach
Individual Behavior (Micro-Laws)
Learn
Societys History
Predict (Simulate)
Societys Future
21
Testing Simulations
22
Prediction
23
Prediction
24
Future Work
  • Test Other Predictions
  • e.g., membership in a particular group
  • Learn from Other Real Data
  • e.g., emails and blogs

25
Questions?
Write a Comment
User Comments (0)
About PowerShow.com