1

Computataional Aspects of Social

Machines

Jie Tang

Tsinghua University

2

Social Networks • >900 million users

• the 3rd largest “Country” in the world

• More visitors than Google

• More than 5 billion images

• 2009, 2 billion tweets per quarter

• 2010, 4 billion tweets per quarter

• 2011, tweets per quarter

• >500 million users

• 2012, users, 300% yearly increase

• Pinterest, with a traffic higher than Twitter and Google

25 billion

300 million

3

A Trillion Dollar Opportunity

Social networks already become a bridge to connect our

daily physical life and the virtual web space

On2Off [1]

[1] Online to Offline is trillion dollar business

http://techcrunch.com/2010/08/07/why-online2offline-commerce-is-a-trillion-dollar-opportunity/

4

What are the fundamentally new

things in social networks?

+ +

+

-

-

-

- +

+

?

? ?

?

? ? ?

?

hyperlinks between web pages

Examples:

Google search (information retrieval)

Web 1.0

5

+

+

+ -

-

-

+ ?

?

? ?

?

?

Collaborative Web

(1)personalized learning

(2)collaborative filtering

What are the fundamentally new

things in social networks?

6

Social Web

(1) interactions

(2) information diffusion

+

+

+ -

-

-

+ ?

?

? ?

?

?

Interactions

1. Influence

2. Collective

intelligence

influence

What are the fundamentally new

things in social networks?

7

Iceberg Model for Social Network

User behaviors

Network

structure

Social Tie

Influence

Collective

Intelligence

Information

Diffusion

8 KDD 2010, PKDD 2011 (Best Paper Runnerup), WSDM 2012, DMKD

Social Ties Analysis

? Family

Friend

Collaborate with

John Hopcroft, Jon Kleinberg (Cornell), Tiancheng Lou (Google)

9

Example: finding boss in email networks (PKDD’11, Best Paper Runnerup)

CEO

Employee

How to

infer Manager

Enterprise email network

User interactions may form implicit groups

10

Networks

• Epinions a network of product reviewers: 131,828 nodes (users)

and 841,372 edges

– trust relationships between users

• Slashdot: 82,144 users and 59,202 edges

– “friend” relationships between users

• Mobile: 107 mobile users and 5,436 edges

– to infer friendships between users

• Coauthor: 815,946 authors and 2,792,833 coauthor relationships

– to infer advisor-advisee relationships between coauthors

• Enron: 151 Enron employees and 3572 edges

– to infer manager-subordinate relationships between users.

Undirected network

Directed network

11 KDD 2009, KDD 2011, DMKD

Social Influence Analysis

Collaborate with

Jimeng Sun (IBM TJ Watson), Jiawei Han and Chi Wang (UIUC)

12

Ada

Frank

Eve David

Carol

Bob

George

2

1

14

2

2 33

Marketer

Alice

Opinion Leader

Find K nodes (users) in a social network that could maximize the

spread of influence (Domingos, 01; Richardson, 02; Kempe, 03)

Social influence

Who are the

opinion leaders

in a community?

13

Ada

Frank

Eve David

Carol

Bob

George

2

1

14

2

2 33

Marketer

Alice

Opinion Leader

Find K nodes (users) in a social network that could maximize the

spread of influence (Domingos, 01; Richardson, 02; Kempe, 03)

Who are the

opinion leaders

in a community?

Challenge: How to quantify the strength of social

influence between users?

Social influence

14

Topic-level Social Influence Analysis

Ada

Frank

Eve David

Carol

Bob

George

Input: coauthor network

Ada

Frank

Eve David

Carol

George

Social influence anlaysis

θi1=.5

θi2=.5

Topic

distributiong(v1,y1,z)θi1

θi2

Topic

distribution

Node factor function

f (yi,yj, z)

Edge factor function

rz

az

Output: topic-based social influences

Topic 1: Data mining

Topic 2: Database

Topics:

Bob

Output

Ada

Frank

Eve

BobGeorge

Topic 1: Data mining

Ada

Frank

Eve David

George

Topic 2: Database

. . .

2

1

14

2

2 33

Several key challenges:

• How to differentiate the social influences from different topics?

• How to incorporate different information (e.g., topic distribution

and network structure) into a unified model？

• How to estimate the model on real-large networks?

[1] J. Tang, J. Sun, C.Wang, and Z. Yang. Social influence analysis in largescale networks. In Proceedings of the 15th ACM SIGKDD

International Conference on Knowledge Discovery and Data Mining (SIGKDD’09), pages 807–816, 2009.

15

Emotion/Action Prediction: Can we infer users’ emotions and social action?

Collaborate with

Jinghai Rao (Nokia), Jimeng Sun (IBM TJ Watson), Yuan Zhang (MIT)

KDD 2010, ICDM 2010, ACM TKDD, IEEE TAC

16

It's an emotional world we live in

Emotion stimulates the mind 3000 times quicker

than rational thought!!!

It's an emotional world we live in!

Six degree vs. Three degree [Nature; BMJ]

17

“Happy” System

Can we predict users’

emotion?

18

Observations

Location correlation

(Red-happy)

Activity correlation

KO

?

?

?

?

?

GYM

Dorm

The Old

Summer

Palace

class

room

19

Observations (cont.)

Social correlation

(a) Social correlation (a) Implicit groups by emotions

Temporal correlation

20

MoodCast: Dynamic Continuous Factor

Graph Model

Jennifer

Happy

Happy

location

Neutral

Neutral

call

sms

Mike

Allen

MikeAllen

Jennifer today

Jennifer

yesterday

?

Jennifer

tomorrow

MoodCast

Predict

Attributes f(.)

Temporal

correlation h(.)

Social correlation g(.)

Our solution

1. We directly define continuous feature function;

2. Use Metropolis-Hasting algorithm to learn the factor graph model.

21

John

Time t

John

Time t+1

Action Prediction：

Will John post a tweet on “Haiti Earthquake”?

Personal attributes:

1. Always watch news

2. Enjoy sports

3. ….

Influence 1

Action bias 4

Dependence 2

Social Action Modeling and Prediction

Correlation 3

22

“Social Machine”

• Deploy a “machine” on Weibo.com, the largest “Twitter” in China;

• Act as a person by auto follow/retweet/reply;

• Attracted thousands of fans.

23

Summaries

• Social network brings a trillion dollar

opportunity

• Computational models – Social tie analysis

– Social influence analysis

– Emotion and Action prediction

24

Representative Publications • Tiancheng Lou, Jie Tang, John Hopcroft, Zhanpeng Fang, Xiaowen Ding. Learning to Predict

Reciprocity and Triadic Closure in Social Networks. ACM TKDD, 2012.

• Honglei Zhuang, Jie Tang, Wenbin Tang, Tiancheng Lou, Alvin Chin, and Xia Wang. Actively

Learning to Infer Social Ties. DMKD, 2012, Vol 25, Issue 2, pages 270-297.

• Lu Liu, Jie Tang, Jiawei Han, and Shiqiang Yang. Learning Influence from Heterogeneous Social

Networks. DMKD, 2012, Vol 25, Issue 3, pages 511-544.

• Jie Tang, Yuan Zhang, Jimeng Sun, Jinghai Rao, Wenjing Yu, Yiran Chen, and ACM Fong.

Quantitative Study of Individual Emotional States in Social Networks. IEEE TAC, 2012, Vol 3, Issue 2,

Pages 132-144. (Spotlight Paper)

• Jie Tang, Tiancheng Lou, and Jon Kleinberg. Inferring Social Ties across Heterogeneous Networks.

WSDM’12. pp. 743-752.

• Chenhao Tan, Lillian Lee, Jie Tang, Long Jiang, Ming Zhou, and Ping Li. User-level sentiment

analysis incorporating social networks. SIGKDD'11. pp. 1397-1405.

• Chenhao Tan, Jie Tang, Jimeng Sun, Quan Lin, and Fengjiao Wang. Social Action Tracking via Noise

Tolerant Time-varying Factor Graphs. KDD’10.

• Chi Wang, Jiawei Han, Yuntao Jia, Jie Tang, Duo Zhang, Yintao Yu, Jingyi Guo. Mining Advisor-

Advisee Relationships from Research Publication Networks. KDD’10.

• Jie Tang, Jimeng Sun, Chi Wang, and Zi Yang. Social Influence Analysis in Large-scale Networks.

KDD'09.

• Jie Tang, Jing Zhang, Limin Yao, Juanzi Li, Li Zhang, and Zhong Su. ArnetMiner: Extraction and

Mining of Academic Social Networks. KDD’08.

25

HP: http://keg.cs.tsinghua.edu.cn/jietang/

System: http://arnetminer.org

Thanks!