Transcript Social Action Tracking via Noise Tolerant Time-varying Factor Graphs 1Chenhao Tan, 1Jie Tang, 2Jimeng Sun, 3Quan Lin, 4Fengjiao Wang 1Department of Computer Science and Technology,

Social Action Tracking via Noise Tolerant
Time-varying Factor Graphs
1Chenhao
Tan, 1Jie Tang, 2Jimeng Sun, 3Quan Lin, 4Fengjiao Wang
1Department
of Computer Science and Technology, Tsinghua University, China
2IBM TJ Watson Research Center, USA
3Huazhong University of Science and Technology, China
4Beijing University of Aeronautics and Astronautics, China
1
Motivation
• 500 million users
• the 3rd largest “Country” in the world
• More visitors than Google
• Action: Update statues, create event
• More than 4 billion images
•Action: Add tags, Add favorites
• 2009, 2 billion tweets per quarter
• 2010, 4 billion tweets per quarter
•Action: Post tweets, Retweet
3
User Action in Social Networks
Twitter
Post tweets
on “Haiti
Earthquake”
4
Flickr
Add photo to
her favorites
Arnetminer
Publish in
KDD
Conference
User Action in Social Networks
Questions:
-What factors influence you to add a photo
into your favorite list?
- If you post a tweet on “Haiti Earthquake”,
will your friends retweet it or reply?
Challenge:
- How to track and model users’ actions?
- How to predict users’ actions over time?
5
Complex Factors
1
Influence
Time t+1
Time t
John
2
6
Action Prediction：
3 on
Will John post a tweet
Correlation
“Haiti Earthquake”?
Dependence
John
Attributes:
4 Personal
1. Always
watch news
attributes
2. Enjoy sports
3. ….
Problem formulation
Gt =(Vt, Et, Xt, Yt)
Actions at time t
Nodes at time t
Edges at time t
Attribute matrix at time t
Input:
Gt =(Vt, Et, Xt, Yt)
t = 1,2,…T
7
Output:
F: f(Gt) ->Yt
NTT-FGM Model
Influence
Correlation
Continuous latent action state
Dependence
Action
8
Personal attributes
Personal attributes
Model Instantiation
How to estimate the parameters?
9
Model Learning
Extremely time costing!!
Our solution: distributed learning (MPI)
10
Experiment
• Data Set
Action
Nodes
#Edges
Action Stats
Twitter
Post tweets on
“Haiti Earthquake”
7,521
304,275
730,568
Flickr
Add photos into
favorite list
8,721
485,253
485,253
Arnetminer
Issue publications
on KDD
2,062
34,986
2,960
• Baseline
– SVM
– wvRN (Macskassy, 2003)
• Evaluation Measure:
Precision, Recall, F1-Measure
12
Performance Analysis
13
Factor Contribution Analysis
• NTT-FGM:
Our model
• NTT-FGM-I: Our model ignoring influence
• NTT-FGM-CI: Our model ignoring influence and correlation
14
Efficiency Performance
15
Conclusion
• Formally formulate the problem of social
action tracking
• Propose a unified model: NTT-FGM to
simultaneously model various factors
• Present an efficient learning algorithm and
develop a distributed implementation
• Validate the proposed approach on three
different data sets, and our model achieves
a better performance
16
Thank you！
QA?
Data & Code: http://arnetminer.org/stnt
Welcome to our poster!
17