Download presentation
Presentation is loading. Please wait.
Published byAliza Williamson Modified over 9 years ago
1
Predicting Emerging Social Conventions in Online Social Networks Farshad Kooti * Winter Mason † Krishna Gummadi * Meeyoung Cha ‡ MPI-SWS * Stevens Institute of Technology † KAIST ‡ CIKM 2012
2
Prediction of Emerging Social Conventions in OSNs- Farshad Kooti Imperial Metric 2
3
Linguistic conventions Prediction of Emerging Social Conventions in OSNs- Farshad Kooti Hey Aloha How’s it going Hello 3
4
The retweeting convention Quoting another user while citing the original author 4 Prediction of Emerging Social Conventions in OSNs- Farshad Kooti Bob Alice RT @Bob: CIKM started RT @Bob: CIKM started RT @Bob: CIKM started RT @Bob: CIKM started
5
Why retweeting convention? o Information-sharing channels are explicit in Twitter o Specific to Twitter: exposures within the community o Contained in Twitter, hence capturing all usages Prediction of Emerging Social Conventions in OSNs- Farshad Kooti5
6
Twitter dataset o Used near-complete data from 03-2006 to 09-2009 -54 million users -1.9 billion tweets -1.7 billion follow links o Follow links are a snapshot of the network in 2009 Prediction of Emerging Social Conventions in OSNs- Farshad Kooti6
7
The retweeting variations o Searched for syntax token @username o “Adopter” refers to a user using the variation at least once Variation# of adopters# of retweets RT1,836 K53,221 K via751 K5367 K Retweeting50 K296 K Retweet36 K110 K HT8 K22 K R/T5 K28 K 3 K18 K Total2,059 K59,065 K Prediction of Emerging Social Conventions in OSNs- Farshad Kooti7
8
Our study of retweeting convention 1.Characterizing the emergence [ICWSM’12, best paper award] 2.Predicting the adoption process [this work, CIKM 2012] Prediction of Emerging Social Conventions in OSNs- Farshad Kooti8
9
Defining prediction problem Suppose we are given a social network with records of users, their interactions, and times of adoptions. However, information about which variation was adopted by user u at time t is hidden. How reliably we can infer that user u has adopted variation v at time t? Prediction of Emerging Social Conventions in OSNs- Farshad Kooti9
10
10 RT or via or...? RT @john: tweet tweet (RT @joe) via @jane: tweet 2,053 TWEETS 1,738 FOLLOWING 1,581 FOLLOWERS Bob
11
Motivation & Problem Features impacting adoption Predictive power & results
12
Feature categories Prediction of Emerging Social Conventions in OSNs- Farshad Kooti12 Personal Social Global
13
feature: # of followers Prediction of Emerging Social Conventions in OSNs- Farshad Kooti13 Personal
14
features Prediction of Emerging Social Conventions in OSNs- Farshad Kooti14 # of exposures # of adopter friends Social
15
feature: # of adopter friends Prediction of Emerging Social Conventions in OSNs- Farshad Kooti15 Social
16
feature: adoption date Prediction of Emerging Social Conventions in OSNs- Farshad Kooti16 Global
17
All the considered features – # of followers and friends, # of posted tweets and URLs, join date, geo-location – # of exposures, # of adopter friends – Time of adoption Prediction of Emerging Social Conventions in OSNs- Farshad Kooti17 Global Social Personal
18
Motivation & Problem Features impacting adoption Predictive power & results
19
Measuring the predictive power of features o We calculate Information Gain (IG) of each feature, which shows the predictive power o IG: change in entropy (measure of uncertainty) because of the given feature o IG(Variation, feat.) = H(Variation) - H(Variation|feat.) Prediction of Emerging Social Conventions in OSNs- Farshad Kooti19
20
Predictive power of features: results Prediction of Emerging Social Conventions in OSNs- Farshad Kooti20 RankFeatureType 1DateGlobal 2# of exposures to RTSocial 3# of posted URLsPersonal 4# of exposures to viaSocial 5Join date of adopterPersonal 6# of posted tweetsPersonal 7# of RT adopter friendsSocial Findings: # of exposures has more predictive power than # of adopter friends Geography is not important
21
Prediction methodology o Using different ML classifiers: Bayesian models, boosting, decision trees, etc. – Bagging yields the best result o Feature selection techniques to find best subset of features (excluded 8 features) Prediction of Emerging Social Conventions in OSNs- Farshad Kooti21
22
Prediction accuracy VariationAccuracyPrecisionRecall RT71.272.868.1 via72.652.166.6 Retweeting98.043.190.5 Retweet98.534.380.1 HT99.750.584.9 R/T99.819.081.5 recycle icon99.935.982.3 Weighted average72.665.769.8 Prediction of Emerging Social Conventions in OSNs- Farshad Kooti22
23
Dealing with unbalanced classes o Problem: – Most of the adoptions (68%) are RT – A simple classifier of always predicting the most used variation performs good o Solution: – Take the same number of cases from two groups (baseline: 50%) Prediction of Emerging Social Conventions in OSNs- Farshad Kooti23
24
Prediction accuracy from balanced data VariationAccuracyPrecisionRecall RT61.360.763.1 via60.760.660.1 Retweeting59.158.961.8 Retweet56.956.6 HT82.382.881.5 R/T77.377.077.2 recycle icon81.583.180.2 Weighted average61.060.761.5 Prediction of Emerging Social Conventions in OSNs- Farshad Kooti24
25
Stronger definitions Prediction of Emerging Social Conventions in OSNs- Farshad Kooti25
26
Summary o Predicting adoption of social conventions o Investigated impact of various factors o Global feature trumps social and personal features o The number of exposures had more predictive power than number of adopter friends o Using the features from network is not enough for a prediction with high accuracy Prediction of Emerging Social Conventions in OSNs- Farshad Kooti26
27
Thank you! Prediction of Emerging Social Conventions in OSNs- Farshad Kooti27
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.