Social/Network/Analysis mohamed.bouguessa@uqo.ca/ 1
Web/today 2
Web/today/–/Diverse/applications 3
Web/today/–/Millions/of/users 4
Web/today/–/Rich/content 5
Web/today/–/Highly/dynamic/ 6
Web/today/–/Traces/of/activity 7
Web/today/–/Rich/interactions Rich/interactions/ between/users/ and/content/ 8
Web/today/–/Interaction/networks Rich interactions between users and content Modeled as interaction network 9 9
Six/degrees/of/separation We#can#all#be#connected#through#a#series#of#six#contacts# appeals#to#me.#It#makes#the#world#seem#less#brutal,#and# more#warm#and#more#friendly.## 10
Six/degrees/of/separation 11
Testing/the/smallGworld/hypothesis MSN Messenger Network of who talks to Average path length is whom on MSN Messenger: 6.6 240M nodes, 1.3 billion 90% of nodes is edges reachable <8 steps 12
Why/study/networks? • Build/understanding/and/theory:/ – How#users#create#content#and#interact#with#it#and# among#themselves?# • Build/better/onGline/applications:/ – How#to#design#better#services#and#algorithms?# 13
Social/Networks/Analysis • #A# social/network/ is#a#social#structure#of#people,#related# (directly#or#indirectly)#to#each#other#through#a#common# relation#or#interest.# • /Social/network/analysis/(SNA)/ is#the#study#of#social# networks#to#understand#their#structure#and#behavior.# 14
Social/Networks • Social#network:#relationship#among# interacting#units.# 15
Social/Networks Interacting unites: Actors / nodes discrete individual, corporate, or collective social units 16
Social/Networks Relational ties between actors are channels to transfer, Relations, exchange or linkages or ties flow of resources. 17
Social/Networks • Social#network#representation# – Adjacency#matrix##(socioGmatrix)# – Graph#(SocioGgraph) # [ ] 1 2 3 4 5 6 7 8 9 0 1 1 0 1 1 0 0 0 1 & # & # $ ! $ ! 1 0 0 0 1 0 0 0 1 2 $ ! $ ! 1 0 0 0 0 1 0 0 0 3 $ ! $ ! $ ! $ ! 4 0 0 0 0 1 0 0 0 1 $ ! $ ! $ ! 5 $ 1 1 0 1 0 0 1 0 1 ! $ ! $ ! 6 1 0 1 0 0 0 0 0 0 $ ! $ ! $ ! 7 $ ! 0 0 0 0 1 0 0 0 1 $ ! $ ! 8 $ ! 0 0 0 0 0 0 0 0 0 $ ! $ ! 9 $ ! 0 1 0 1 1 0 1 0 0 % " % " 18
Key/Drivers/for/CS/Research/in/SNA • #Computer#Science#has#created#the#cyber#infrastructure#for# –#Social#Interaction# –#Knowledge#Exchange# –#Knowledge#Discovery# • #Ability#to#capture# –#different#about#various#types#of#social#interactions# –#at#a#very#fine#granularity# –#with#practically#no#reporting#bias# Data/mining/techniques/can/be/used/for/building/ descriptive/and/predictive/models/of/social/interactions/ 19
SNA/Techniques Prominent/problems/ • Social#network#extraction/construction# • Identifying#prominent/trusted/expert#actors# • Identifying#Spammers### • Discovering#communities#in#social#networks# • Link#prediction# • Approximating#large#social#networks# • Evolution#of#social#networks# 20
Social/Network/Extraction • #Mining#a#social#network#from#data#sources# • #Recent#research#suggest#that#there#are#three#sources#of#social# network#data#on#the#web# • #Content#available#on#web#pages#(e.g.#user#homepages,#message# threads#etc.)# • #User#interaction#logs#(e.g.#email#and#messenger#chat#logs)# • #Social#interaction#information#provided#by#users#(e.g.#social# network#service#websites#such#as#Orkut,#Friendster#and#MySpace)# P P P r r r o o o f f f i i i l l l e e e _ _ _ 1 1 1 P P P r r r o o o f f f i i i l l l e e e _ _ _ 3 3 3 P P P r r r o o o f f f i i i l l l e e e _ _ _ 2 2 2 W W e e b b D D o o c c u u m m e e n n t t s s P P P r r r o o o f f f i i i l l l e e e _ _ _ 5 5 5 P P P r r r o o o f f f i i i l l l e e e _ _ _ 4 4 4 C C C o o o m m m m m m u u u n n n i i i c c c a a a t t t i i i o o o n n n L L L o o o g g g s s s A A A c c c t t t o o o r r r p p p r r r o o o f f f i i i l l l e e e s s s o o o n n n a a a S S S o o o c c c i i i a a a l l l N N N e e e t t t w w w o o o r r r k k k S S S e e e r r r v v v i i i c c c e e e 21
Social/Network/Extraction • Extracting#a#social#network# – Asking#people#about#their#relations# – Tracking#their#contacts#(emails,#phone#call,# visits,#etc.)#such#as#Enron#project# – Mining#their#contextual#data#(papers,# interviews,#resumes,#news,##biographies,# citations,#references,#web#pages,#blogs,# portfolios,#etc.)# ! #Learning#social#network# 22
Learning/Social/Networks • Learning##social#network#from#text# – Descriptive#vs.#Predictive#model# – We#only#predict#the#possible#relations# between#the#actors# 23
Learning/Social/Networks Usually,#we#can#reach#documents#by# knowing#people…# # 24
Learning/Social/Networks …and#directly#or#indirectly#we#will#know# other#documents#by#(or#about)#other#people# through#these#documents…# # # 25
Learning/Social/Networks …and#very#soon#we#will#have#a#social#network# including#some#individuals#who#have#been# connected#to#each#other#via#some#similar# contents.# # 26
SNA/Techniques Prominent/problems/ • Social#network#extraction/construction# • Identifying/prominent/trusted/expert/actors/ • Identifying#Spammers### • Discovering#communities#in#social#networks# • Link#prediction# • Approximating#large#social#networks# • Evolution#of#social#networks# 27
Identifying/prominent/expert/actors/ in/social/networks/ Link/Analysis/Technique/ • HITS/ • PageRank/ 28
Hubs/and/Authorities • #Being#Authority#depends#upon#inGedges;#an#authority#has#a#large# number#of#edges#pointing#towards#it.# • #Being#a#Hub#depends#upon#outGedges;#a#hub#links#to#a#large# number#of#nodes.# • #Notice#that#the#definition#of#hubs#and#authorities#is#circular.# • #Nodes#can#be#both#hubs#and#authorities#at#the#same#time# 29
Hubs/and/Authorities h ( p ) a ( q ) a ( p ) h ( q ) ∑ ∑ = = p q q p → → 30
Google’s/PageRank • # The# PageRank# assumption# is# that# a# node# transfers# its# PageRank#values#evenly#to#all#the#nodes#it#connects#to.## • ##A#node#has#high#rank#if#the#sum#of#the#ranks#of#its#inGlinks#is# high.### • #This#covers#both#the#case#where#a#node#has#many#inGlinks# and#that#where#a#node#has#a#few#highly#ranked#inGlinks.# 31
Google’s/PageRank How/is/PageRank/calculated?/ PR ( T 1 ) PR ( Tn ) & # PR ( A ) ( 1 d ) d * = − + + + $ ! C ( T 1 ) C ( Tn ) % " C(Ti):#the#number#of#OutGlinks#of#the#page/node#Ti# That's#the#equation#that#calculates#a#page's#PageRank.#It's#the# original#one#that#was#published#when#PageRank#was#being# developed,#and#it#is#probable#that#Google#uses#a#variation#of#it# but#they#aren't#telling#us#what#it#is.#It#doesn't#matter#though,#as# this#equation#is#good#enough.## 32
Google’s/PageRank PR(A)#=#PR(B)#=#PR(C)#=PR(D)#=1# PR(A)#>#PR(B)#>#PR(C)#>#PR(D)## 33
Google’s/PageRank A A B D E B D C C A A E B D B D E C C 34
Yahoo!/Answers/:/Identifying/the/expert/ User User_x Votes Question User_y Answer_1 User_1 Answer_2 User_2 User_z Answer_n User_n • #Identifying#the#true#experts#among#Yahoo#Answers#participants# • #Keep#track#of#users#who#consistently#provide#good#answers#for#particular#topics# • #Provide#incentives#for#experts#to#stay#on#Yahoo!#Answers#in#order#to#improve# service# 35
Yahoo!/Answers 36
Question/Life/Cycle 37 37
Yahoo!/Answers Example#of#interactions#between#askers#and#best#answerers## 5 2 3 3 1 2 3 1 2 1 2 1 4 1 1 4 4 Users who usually only ask questions Users who usually only answer questions Users who help each other How$to$es(mate$the$authority$degree$for$each$user? 38
PageRank? Example :#The#category#of#“Programming” / • #User# B #answers#user# A ’s#questions,#which#are#about#Java;# • #User# C #answers# B ’s#questions,#which#are#about#PHP;### JAVA PHP A B C " #Is#it#possible#to#state#that# C #is#more#expert#than# B ?## • #No,#because:# B #and# C #have#different#expertise.## 39
HITS? • #The#HITS#algorithm#assigns#high#authority#scores#to# nodes#1,#2,#10,#11#and#12,#but#a#nearGzero#authority#score# to#node## 40
HITS? The#HITS#algorithm#will#allocate#high#authority#scores#to#the#nodes#N9–N15,# • while#giving#zero#authority#score#to#node#N1.## The#reason#for#this#is#quite#similar#to#example#1.## • Specifically,#the#fact#that#node#N8#points#to#many#nodes#contribute#to#increase# • its#hub#score.#Hence,#causing#the#nodes#N9–N15#to#receive#higher#authority# scores.#However,#intuition#suggests#that#node#N1#is#the#most#authoritative# since#it#represents#an#answerer#with#a#large#number#of#best#answers.# 41
Recommend
More recommend
