0.25
0.5
0.75
1.25
1.5
1.75
2
Learning to Infer Social Ties in Large Network
Published on Oct 03, 20113315 Views
In online social networks, most relationships are lack of meaning labels (e.g., "colleague" and "intimate friends"), simply because users do not take the time to label them. An interesting question
Related categories
Chapter list
Learning to Infer Social Ties in Large Networks00:00
Real social networks are complex...00:08
Even complex than we imagined!02:05
Example: Mobile network02:59
Example: Coauthor networks03:22
Challenges (1)03:54
Challenges (2)04:03
Problem Formulation (1)04:15
Problem Formulation (2)04:26
Problem Formulation (3)04:28
Problem Formulation (4)04:38
Problem Formulation (5)04:41
Problem Formulation (6)04:48
Basic Idea04:58
Partially Labeled Pairwise Factor Graph Model (PLP-FGM) (1)05:34
Partially Labeled Pairwise Factor Graph Model (PLP-FGM) (2)05:47
Partially Labeled Pairwise Factor Graph Model (PLP-FGM) (3)06:06
Partially Labeled Pairwise Factor Graph Model (PLP-FGM) (4)06:28
Partially Labeled Pairwise Factor Graph Model (PLP-FGM) (5)07:05
Partially Labeled Pairwise Factor Graph Model (PLP-FGM) (6)07:43
Solutions (con’t)07:53
Learning Algorithm08:23
Challenges (3)09:23
Distributed Learning09:40
Data Sets11:32
Baselines12:02
Performance Analysis (1)12:52
Performance Analysis (2)13:02
Factor Contribution Analysis (Email) (1)13:22
Factor Contribution Analysis (Email) (2)15:16
Distributed Learning Performance15:22
System on http://arnetminer.org15:54
Conclusion16:23
Future work16:43
Thank you!17:33