1. lidong.wu@utdallas.edu Computational Social Networks --computational data networks Weili Wu Ding-Zhu Du University of Texas at Dallas

2. Goal This course contains advanced topics in one of computational social networks. The social network is one of data networks. The goal of this course has two folds: (a) Let students learn techniques in design and analysis of algorithms, especially approximation algorithms. (b) Lead students to frontier of research in computational data networks. 2

3. Reference books 3

4. Not “Prerequisites”

5. lidong.wu@utdallas.edu Upcoming Springer Book: Optimal Social Influence Wen Xu, Weili Wu University of Texas at Dallas

6. lidong.wu@utdallas.edu Lecture 1-1 What is a Social Network? Ding-Zhu Du University of Texas at Dallas

7. Outline  Social Network  Online Social Networks  Community Structure  Rumor Blocking  Power Law 7

8. Web definition: A network consists of two or more nodes that are linked in order to share resources. What is a Network? 8

9. 2

10. What is Social Network? Wikipedia Definition: Social Structure Nodes: Social actors (individuals or organizations) Links: Social relations 10

11. Example 1: Friendship Network Nodes: all persons in the world A link exists between two persons if they know each other. 11

12. Milgram (1967) The experiment: Random people from Nebraska were to send a letter (via intermediaries) to a stock broker in Boston. Could only send to someone with whom they know. Six links were needed. Stanley Milgram (1933-1984) Property of Friendship Six Degrees of Separation 12

13. Chinese Observation 八竿子打不着 形容二者之间关系疏远或毫无关联。 “ 竿 ” 也 作 “ 杆 ” 。 13

14. Family Friend Family Friend Supervise Friend Roommate Friend 14 Lidong Wu

15. “The small world network is a type of mathematical graph in which most nodes are not neighbors of one another, but most nodes can be reached from every other by a small number of hops or steps.”

16. Example 2: Coauthorship Network Nodes: all publication authors A link exists between two authors if they are coauthors in a publication. 16

17. Erdős number: is the collaboration distance with mathematician Paul Erd ő s. What is your Erdős number? Erdös number 0 --- 1 person Erdös number 1 --- 504 people Erdös number 2 --- 6593 people Erdös number 3 --- 33605 people Erdös number 4 --- 83642 people Erdös number 5 --- 87760 people Erdös number 6 --- 40014 people Erdös number 7 --- 11591 people Erdös number 8 --- 3146 people Erdös number 9 --- 819 people Erdös number 10 --- 244 people Erdös number 11 --- 68 people Erdös number 12 --- 23 people Erdös number 13 --- 5 people * Two persons are linked if they are coauthors of an article. Coauthorship Network is a Small World Network 17 Distribution in Dec.2010

18. My Erdős number is 2. 18

19. Nodes: all cities with an airport. A link exists between two cities if there exists a direct flight between them. Example 3: Flight Map Is a Small World Network 19

20. Find a cheap ticket between two given cities. It is a shortest path problem in a social network. Need to add connection information to network. Search Cheap Ticket 20 There are about 28,537 commercial flights in the sky in the U.S. on any given day.

21. Network Construction AA123 AA456 AA789 Dallas Chicago 21

22. Network Construction 22 Dallas 8am 9am 1pm 9am 3pm 8am

23. Network construction Dallas 8am 9am 1pm 9am 3pm 8am 23

24. Outline  Social Network  Online Social Networks  Community Structure  Rumor Blocking  Power Law 24

25. Social Network is online in Internet Facebook: friendship linkedIn: friendship ResearchGate: coauthorship 25

26. Online Social Networks (OSN) Social influence occurs when one's emotions, opinions, or behaviors are affected by others. Although social influence is possible in the workplace, universities, communities, it is most popular online.

27. Internet provides a platform to record and to develop social networks 27

28. What Are OSN Used For? 28

29. Candidates (left to right) ： Ken Livingstone, Boris Johnson and Brian Paddick. Political Election for Mayor of London (2012) Usage Example http://www.telegraph.co.uk /technology/news/923907 7/Twitter-data-predicts- Boris-Johnson-victory.html 29

30. Prediction of Boris Johnson Victory 30

31. How to Predict? Analysis posts on Facebook and Twitter: “Sentiment Analysis”. Find 7% more positive sentiment towards Mr. Johnson than Mr. Livingstone. Predict 54% of the vote for Mr. Johnson. Google Insights, tracking web trends, Almost five times more searches for “Boris Johnson” than for “Ken Livingstone” via google.co.uk. Of the total number of web searches for both candidates, 60% were for “Boris Johnson”. 31

32. Outline  Social Network  Online Social Networks  Community Structure  Rumor Blocking  Power Law 32

33. Question 1? Does Six Degrees of Separation imply six degrees of influence? 33

34. Three Degrees of Influence in friendship network 34

35. Three Degrees of Influence In Book Connected by Nicholas A. Christakis and James H. Fowler. Nicholas A. ChristakisJames H. Fowler

36. Three Three Degrees of Influence The influence of actions ripples through networks 3 hops (to and from your friends’ friends’ friends). 36

37. I am happy! 37

38. Question 2? How to explain Six Degrees of Separation and Three Degrees of Influence? 38

39. Community People in a same community share common interests in - clothes, music, beliefs, movies, food, etc. Influence each other strongly. 39

40. * same color, same community Community without overlap Community with overlap Community Structure 40

41. two nodes can reach each other in three steps. A few of tied key persons: C, D Member A reaches Member B via A-C-D-B Community Structure 41 In the same community,

42. Two nodes may have distance more than three. Community Structure 42 For different communities,

43. Community Structure Two nodes can reach each other by at most six steps. A C B 43 For two overlapping communities,

44. Outline  Social Network  Online Social Networks  Community Structure  Rumor Blocking  Power Law 44

45. 6/9/201645 When misinformation or rumor spreads in social networks, what will happen?

46. A misinformation said that the president of Syria is dead, and it hit the twitter greatly and was circulated fast among the population, leading to a sharp, quick increase in the price of oil. http://news.yahoo.com/blogs/technology-blog/twitter-rumor- leads-sharp-increase-price-oil-173027289.html 6/9/201646

47. In August, 2012, thousands of people in Ghazni province left their houses in the middle of the night in panic after the rumor of earthquake. http://www.pajhwok.com/en/2012/08/20/quak e-rumour-sends-thousands-ghazni-streets 6/9/201647

48. 6/9/201648 People in a same community share common interests in - clothes, music, beliefs, movies, food, etc. Influence each other strongly.

49. Rumor Blocking Problem 67 5 1 3 4 2 8 9 10 11 12 13 14 Yellow nodes are bridge ends. 6/9/201649

50. Example 1 3 4 5 2 6 1 is a rumor, 6 is a protector. Step 1: 1--2,3; 6--2,4. 2 and 4 are protected, 3 is infected. 6/9/201650 rumor protector

51. 1 3 5 2 4 6 Step 2: 4--5. 5 is protected. Example 6/9/201651

52. Least Cost Rumor Blocking Problem (LCRB) Bridge ends:  form a vertex set;  belong to neigborhood communities of rumor community;  each can be reached from the rumors before others in its own community. C0 C2 C1 Red node is a rumor; Yellow nodes are bridge ends. 6/9/201652

53. Set Cover Problem 67 5 1 3 4 2 8 9 10 11 12 13 14 Yellow nodes are bridge ends. 6/9/201653

54. Greedy Algorithm 54

55. Outline  Social Network  Online Social Networks  Community Structure  Rumor Blocking  Power Law 55

56. What is Power Law Graph?

57. Less nodes with higher degree and more nodes with lower degree. All peoples are surround leaders. Community Structure 57

58. During the evolution and growth of a network, the great majority of new edges are to nodes with an already high degree. Power Law 58

59. Power law distribution: f(x) ~ x –α Log-log scale: log f(x) ~ –αlog x Power-law distribution 59

60. 60

61. Nodes with high degrees may have “butterfly effect”. Small number Big influence Power Law 61

62. Important Facts on Power-law Many NP-hard network problems are still NP- hard in power-law graphs. While they have no good approximation in general, they have constant-approximation in power-law graphs.

63. References 63

64. THANK YOU!