credibility, identity resolution, and privacy on online social media
TRANSCRIPT
Credibility,IdentityResolution,andPrivacyonOnlineSocialMedia
IEEEInternational Conference onComputing,AnalyticsandSecurity Trends(CAST-2016)
COEPDec20,2016
PonnurangamKumaraguru(“PK”)AssociateProfessor
ACMDistinguishedSpeakerfb/ponnurangam.kumaraguru,@ponguru
WhoamI?
� AssociateProfessor,IIIT-Delhi� Ph.D.fromSchoolofComputerScience,
CarnegieMellonUniversity(CMU)� Researchinterests-SocialComputing,ComputationalSocialScience,ComplexNetworkspertainingtoHumanBehavior,specificallyinthecontextofSecurity&Privacy
� Co-ordinateandmanagePrecog,precog.iiitd.edu.in
� ACMDistinguishedSpeaker
2
TrainingData
� 500Tweetsperevent� UsedCrowdFlower
16
Event Tweets UsersBostonMarathonBlasts(2013) 7,888,374 3,677,531
Typhoon Haiyan /Yolanda(2013) 671,918 368,269
CyclonePhailin (2013) 76,136 34,776
WashingtonNavy yard shootings (2013) 484,609 257,682
Polarvortex cold wave (2014) 143,959 116,141
OklahomaTornadoes (2013) 809,154 542,049
Total 10,074,150 4,996,448
CredibilityModeling
17
Featureset Features (45)
Tweetmeta-dataNumberofsecondssincethetweet;Sourceoftweet(mobile/web/etc);Tweetcontainsgeo-coordinates
Tweetcontent(simple)
Numberofcharacters;Numberofwords;NumberofURLs;Numberofhashtags;Numberofuniquecharacters;Presenceofstocksymbol;Presenceofhappysmiley;Presenceofsadsmiley;Tweetcontains`via';Presenceofcolonsymbol
Tweetcontent(linguistic)
Presenceofswearwords;Presenceofnegativeemotionwords;Presenceofpositiveemotionwords;Presenceofpronouns;Mentionofselfwordsintweet(I;my;mine)
Tweetauthor Numberoffollowers;friends;timesincetheuserifonTwitter;etc.
TweetnetworkNumberofretweets;Numberofmentions;Tweetisareply;Tweetisaretweet
Tweet links WOTscorefortheURL;Ratiooflikes/dislikesforaYouTubevideo
Challenges
24
ProfessionalOpinion
Dating
HeterogeneousOSNs
Personal
DegreeofDetails
QualityanddescriptivepersonalAndprofessionalinformation
LittlepersonalinformationDescriptiveopinions
AttributeEvolution
Time
Informationevolvedononebutnotonother
{jainpari,Bangalore}
RegistrationwithsameinformationonbothOSNs{paridhij,NewDelhi}
GenericIdentityResolution
25
Extractavailable&
discriminativefeatures
CandidateIdentities
IDENTITYSEARCH IDENTITYLINKING
PairwiseComparisons
HeuristicIdentitySearch
26cerc.iiitd.ac.in
Profile
Content
Self-mention
Network Syntactic and Image
Search Linking
If self-identified / returned by
more than one search method
No
Yes
Candidate Identities
name, location,usernamemobile no,
post,friends,
followers
ParidhiJain,Ponnurangam Kumaraguru,andAnupam Joshi.2013.@Iseek‘fb.me’:IdentifyingUsersacrossMultipleOnlineSocialNetworks.InProceedingsofthe22ndInternationalConferenceonWorldWideWeb,WWW’13Companion.ACM,NewYork,NY,USA,1259- 1268.DOI=http://dx.doi.org/10.1145/2487788.2488160[HonorableMentionAward}
29
HowmanyofyouhavepostedmobilenumbersonOnlineSocial
Networks?
Howmanyofyouhaveseenmobilenumbersbeingpostedon
OnlineSocialNetworks?
Datastatistics
� Twitter:12thOctober2012– 20thOctober2013� Facebook:16thNovember2012– 20thApril2013
34
Numbers Category+91 Category0 Categoryvoid Total
Twitter Facebook Twitter Facebook Twitter Facebook Twitter Facebook
MobileNumbers
885 2,191 14,909 8,873 25,566 25,294 41,360 36,358
Userprofiles
1,074 2,663 17,913 9,028 31,149 25,406 49,817 36,588
Takeaways
�OnlineSocialMediaisadifferentbeastintermsofprivacy,identity,andcredibility-Research/technologiesshouldbedeveloped
�Multipleinterestingresearch,engineering,andinnovationwaitingtobedone
� Interestedinhostingstudents– B.Tech.,M.Tech.,Ph.D.
37