1. Understanding Images
1.1. Sentiment
1.1.1. you2015robust(Flickr + Twitter)
1.1.2. xo2014visual (Tumble + Twitter)
1.1.3. jisurvey
1.2. Predict Popularity/Virality
1.2.1. dubey2016deep (Flickr + Reddit)
1.2.2. vallet2015characterizing (Twitter + Youtube)
1.2.3. deza2015understanding (Reddit)
1.2.4. khosla2014makes (Twitter)
1.3. Predict Visual Humour
1.3.1. chandrasekaran2015we (Abstract scenes)
1.4. Aesthetic/Style
1.4.1. zhang2015describing (Flickr)
1.4.2. lu2015deep
1.4.3. hibadullah2015colour
2. Text + Images
2.1. Enriching representation
2.1.1. chencontext (Twitter)
2.2. Sentiment
2.2.1. yu2016 (Sina Weibo)
2.2.2. wang2015sentiment (Flickr + Instagram)
2.2.3. wang2016beyond (Twitter + Stanford ST)
2.3. Sarcasm
2.3.1. schifanella2016detecting (Twitter)
2.4. Joint representation
2.4.1. cha2015multimodal (Twitter)
2.5. Geolocation
2.5.1. ramisa2016breakingnews (BreakingNews)
2.6. Popularity prediction
2.6.1. ramisa2016breakingnews (BreakingNews)
3. Understanding Network
3.1. Causality
3.1.1. mcandrew2016we (Twitter)
3.1.2. charlton2016mood (Twitter)
3.2. Inference
3.2.1. friedland2013copy (Birth + cellphone)
3.3. Diffusion
3.3.1. romero2011differences (Twitter)
3.3.2. spasojevic2015post (Twitter)
3.3.3. gomez2015diffusion (Twitter)
3.3.4. oh2016image (Twitter)
3.4. Seed selection
3.4.1. hung2015social (Twitter)
3.4.2. aral2012identifying (Twitter)
4. Understanding Topics
4.1. Entity/Hashtag
4.1.1. saleiro2016learning (Twitter)
4.1.2. romero2011differences (Twitter)
4.2. Event detection
4.2.1. yilmaz2016multimodal (Twitter)
5. Reasons to retract
5.1. Regrettable content
5.1.1. zhou2016tweet (Twitter)
5.1.2. petrovic2013wish (Twitter)
5.1.3. zhou2015identifying (Twitter)
5.1.4. kawase2013wants (Twitter)
5.2. Pornographic content
5.2.1. Images
5.2.1.1. basilio2010explicit
5.2.1.2. ruiz2005characterizing
5.2.1.3. arentz2004classifying
5.3. Hate speech
5.3.1. burnap2014hate (Twitter)
5.3.2. warner2012detecting
5.4. Plagiarism
5.5. Controversial topic
5.5.1. popescu2010detecting (Twitter)
5.6. Bullying
5.6.1. sanchez2011twitter (Twitter)
5.6.2. galan2014supervised (Twitter)
5.7. Profanity
5.7.1. xiang2012detecting (Twitter)
5.7.2. marmol2014reporting
5.7.3. gupta2014tweetcred (Twitter)
5.7.4. chen2012detecting (Youtube comments)
5.7.5. campbell2013content (Twitter)
5.7.6. xu2010filtering (Youtube comments)
6. Understanding Text
6.1. Sentiment
6.1.1. kumar2015ask (Stanford ST)
6.1.2. balikas2016twise (Twitter)
6.1.3. vosough2016enhanced (Twitter)
6.1.4. yin2016multichannel (Twitter)
6.1.5. thakkar2015approaches (Twitter)
6.1.6. haldenwang2015sentiment (Twitter)
6.1.7. das2016fusion
6.1.8. prusa2015utilizing (Twitter)
6.1.9. kouloumpis2011twitter (Twitter)
6.1.10. kharde2016sentiment (Twitter)
6.1.11. croce2016injecting (Twitter)
6.1.12. pak2010twitter (Twitter)
6.1.13. kuefler2016merging (imdb)
6.1.14. nguyen2016leveraging (Twitter)
6.1.15. tang2015deep (Twitter)
6.2. Target-dependent Sentiment
6.2.1. tang2015target (Twitter)
6.3. Predict geolocation
6.3.1. yuan2013and (Twitter)
6.3.2. cheng2010you (Twitter)
7. Understanding user
7.1. Behaviour/Social Role
7.1.1. fang2015collaborative (Behance)
7.1.2. yuan2013and (Twitter)
7.1.3. zhao2013inferring (Twitter)
7.1.4. spasojevic2015post (Twitter)
7.2. Gender
7.2.1. burger2011discriminating (Twitter)
7.3. Recommendation
7.3.1. geng2015learning (Pinterest)
7.4. Cross-network learning
7.4.1. correa2012whacky (Flickr + Foursquare + ..)
7.5. Political alignment
7.5.1. conover2011predicting (Twitter)
7.6. Location
7.6.1. zhu2015modeling (Twitter)
7.6.2. cheng2010you (Twitter)
7.6.3. yuan2013and (Twitter)