VideoLectures.Net at http://videolectures.net/'>http://videolectures.net/
View the talk in context: http://videolectures.net/'>http://videolectures.net/solomon_lesk...
View the complete seminars: http://videolectures.net/'>http://videolectures.net/solomon/
Speaker: Jure Leskovec, Computer Science Department, Stanford University

License: Creative Commons CC BYNCND 3.0
More information at http://videolectures.net/'>http://videolectures.net/site/about/
More talks at http://videolectures.net/'>http://videolectures.net/

The Web, society, information, cells and brain can all be represented and studied as complex networks of interactions. Nodes in such networks tend to organize into clusters and communities, which represent the fundamental structures for understanding the organization of complex systems. Even though detection of network communities is of significant importance for computer science, sociology and biology, our understanding of the community structure of large networks remains limited.

We study a set of more than 200 large networks with the goal to understand and identify communities in networks. We challenge the conventional view of network community structure and show that it is not exhibited by the large realworld networks. We then present a new conceptual model of network community structure, which reliably captures the overall structure of networks and accurately identifies the overlapping nature of network communities.

0:00 Networks, Communities and the GroundTruth
0:17 Many Data ARE Networks
1:35 Organization of Networks
2:11 Network Communities (1)
2:13 Organization of Networks
2:49 Network Communities (1)
4:15 Communities as Micromarkets
5:22 This Talk: Networks & Communities (1)
6:24 This Talk: Networks & Communities (2)
7:03 This Talk: Networks & Communities (3)
7:51 Part 1: CorePeriphery
8:37 Network Communities (2)
9:54 Network Communities (3)
9:55 Communities: Social Networks
11:31 Community Score
13:14 Network Community Profile (1)
15:02 Network Community Profile (2)
16:40 Network Community Profile (3)
16:44 NCP Plot: Lattices
18:43 NCP Plot: Network Science
19:24 Natural Hypothesis (1)
19:39 NCP Plot: Network Science
19:43 Natural Hypothesis (1)
19:58 Natural Hypothesis (2)
20:02 Natural Hypothesis (1)
20:13 Natural Hypothesis (2)
20:28 Large Networks: Very Different!
21:29 More NCP Plots
22:57 NCP: LiveJournal (n=5m, m=42m) 1
23:04 NCP: LiveJournal (n=5m, m=42m) 2
23:32 Explanation: The Upward Part
24:55 Explanation: Downward part
25:48 What If We Remove Good Clusters?
26:51 Suggested Network Structure (1)
27:14 Suggested Network Structure (2)
27:26 Suggested Network Structure (3)
27:35 Suggested Network Structure (4)
28:08 Suggested Network Structure (5)
28:32 Part 2: Networks & Communities
29:01 Step Back: Community Detection
30:21 GroundTruth